Women's participation in the Australian digital content industry

WOMEN’S PARTICIPATION

IN THE

AUSTRALIAN

DIGITAL CONTENT INDUSTRY

Doctor of Philosophy (PhD)

Anitza A. Geneve

Queensland University of Technology

Professor Karen Nelson

Not to be reproduced without permission or full reference to the author as per Creative Commons Attribution license

‘Women’s Participation in the Australian Digital Content Industry’

Author: Anitza Geneve Year: 2013 P a g e ii of 401

Author: Anitza Geneve Year: 2013 P a g e iii of 401

Keywords

Digital Content Industry (DCI), Information Communication Technology (ICT),

Gender, Human Agency, Interactive Content Creator, Multimedia, New Media, Games,

Critical Realism (CR), Theoretical Scaffold, Multi-Theory Scaffold (MTS), Human

Agency Multi-Theory Scaffold (HAMTS)

List of Abbreviations

ABS Australian Bureau of Statistics AIMIA Australian Interactive Media Industry Association ANZSCO Australian and New Zealand Standard Classification of Occupations AU Australia CR Critical Realism DCI Digital Content Industry DCITA Department of Communications, Information Technology and the Arts EEO Equal Employment Opportunity GDAA Games Developers’ Association of Australia GFC Global Financial Crisis ICT Information and Communications Technology IDT Individual Differences Theory IGDA International Game Developers Association IS Information System IT Information Technology MTS Multi-Theory Scaffold SCCT Social Cognitive Career Theory SCT Social Cognitive Theory ST Structuration Theory STG Social Theory of Gender TAFE Technical and Further Education UK United Kingdom UN United Nations US United States of America VET Vocational Education and Training WIT Women in Technology WoW World of Warcraft

Author: Anitza Geneve Year: 2013 P a g e iv of 401

Abstract

The purpose of the reported research is to investigate the research problem:

Why do women participate in the Australian Digital Content Industry (DCI). The

research investigates two research questions: 1) What are the influences on women’s

participation? and 2) How can we understand these influences? The research problem

arises as a result of a global pattern of women traditionally being under-represented in

the industries related to technology use and development, and the continued under-

representation of women in multimedia and games development production.

The research approach employed an exploratory case study, underpinned by a

Critical Realist ontology, which aimed for empirical, theoretical, and explanatory

insights. Multiple sources of qualitative data were collected over the 2007-2011

period. Primary data included the results of semi-structured interviews with 18 female

interactive content creators employed in the DCI in the city of Brisbane, Australia.

Secondary data sources included industry reports and popular media. Data analysis

employed three frameworks: 1) analytical (based on previous research in the domain);

2) theoretical (employing concepts from four Human Agency theories); and 3)

ontological (Critical Realism; specifically, Bhaskar’s three domains of reality). Each

framework guides the identification of the influences on women’s participation in the

DCI, and provides an approach to understanding these influences.

Findings from the analysis of data using each of the frameworks reveals a range

of influences on women’s participation in the DCI, which include: the characteristics

of the environment (such as cultural practices in the workplace) and the individual

(such as self-efficacy); the events women experience (such as becoming a parent); and

the underlying mechanisms women can harness (such as gaining industry-relevant

skills).

Several models are developed to offer an original approach to understanding

women’s participation in the DCI. In addition, the Multi Theory Scaffold (MTS) model

offers an approach for the use of multiple theories to scaffold inductive research. The

key contribution of this research is its synthesis of the analysis of the empirical data

using the three frameworks to develop the Five Acts of Agency (FAA) model and Acts

of Agency theory. The FAA identifies 10 agent-driven mechanisms that manifest in the

interaction between the environment and the individual. Agent-driven mechanisms

focus on the causal effect of people themselves; that is, the role individuals themselves

play in their participation.

Author: Anitza Geneve Year: 2013 P a g e v of 401

Understanding influences from the perspective of agent-driven mechanisms also

provides a new approach to planning and evaluating strategies that aim to foster

women’s participation in the DCI. Fostering women’s participation in the DCI, in turn,

can also address concerns regarding the general participation of women in other areas

of society. The research contributes to an area where there has been little empirical or

theoretical academic research to date.

Author: Anitza Geneve Year: 2013 P a g e vi of 401

Table of Contents

CHAPTER 1: INTRODUCTION ..................................................................................................... 151.1 Personal Motivation .................................................................................................................. 161.2 Context ...................................................................................................................................... 171.3 Problem Statement, Research Questions and Objectives .......................................................... 191.4 Research Approach ................................................................................................................... 201.5 Research Contribution ............................................................................................................... 231.6 Chapter Summary ...................................................................................................................... 24

CHAPTER 2: WOMEN’S PARTICIPATION IN THE DCI ......................................................... 272.1 The Choice of Literature ........................................................................................................... 272.2 An Overview of the Australian DCI .......................................................................................... 292.3 Current Participation of Women in the DCI .............................................................................. 32

2.3.1 What is Participation ...................................................................................................... 322.3.2 Women’s Inequitable Participation ................................................................................ 332.3.3 Why Participation Matters ............................................................................................. 36

2.4 Influences on Women’s Participation........................................................................................ 382.5 Initial Methodological Considerations ...................................................................................... 432.6 Chapter Summary ...................................................................................................................... 50

CHAPTER 3: WAYS TO UNDERSTAND INFLUENCES ........................................................... 523.1 Framework 1: Analytical model (E, P, I) .................................................................................. 53

3.1.1 Environment (E) Influences ........................................................................................... 563.1.2 Person (P) Influences ..................................................................................................... 583.1.3 Interaction (I) between the Environment (E) and Person (P) ......................................... 593.1.4 Summary ........................................................................................................................ 61

3.2 Framework 2: Theoretical (Human Agency Theory) ................................................................ 613.2.1 The role of theory ........................................................................................................... 623.2.2 Multi-Theory Scaffold (MTS) ........................................................................................ 653.2.3 Human Agency Theory .................................................................................................. 713.2.4 A Human Agency Multi-Theory Scaffold (HAMTS) .................................................... 743.2.5 Summary ........................................................................................................................ 80

3.3 Framework 3: Ontological (Critical Realism) ........................................................................... 813.3.1 A Critical Realist Ontology ............................................................................................ 813.3.2 A Critical Realist Framework ........................................................................................ 833.3.3 Summary ........................................................................................................................ 86

3.4 Chapter Summary ...................................................................................................................... 86

CHAPTER 4: RESEARCH DESIGN .............................................................................................. 894.1 Case Study as a Strategy of Enquiry ......................................................................................... 90

4.1.1 Case study context, boundary and unit of analysis ......................................................... 934.2 Data Collection .......................................................................................................................... 95

4.2.1 Participant Sampling ...................................................................................................... 954.2.2 Data Collection .............................................................................................................. 97

4.3 Data Analysis .......................................................................................................................... 1054.3.1 Danermark et al.’s (2002) Six Stage Model of Explanation ......................................... 1064.3.2 Analysis of Data ........................................................................................................... 112

4.4 Quality ..................................................................................................................................... 1154.4.1 Critical realist criteria ................................................................................................... 1154.4.2 Qualitative criteria ........................................................................................................ 1184.4.3 Case study criteria ........................................................................................................ 1194.4.4 Ethics............................................................................................................................ 119

4.5 Theory Generation ................................................................................................................... 1214.6 Chapter Summary .................................................................................................................... 123

Author: Anitza Geneve Year: 2013 P a g e vii of 401

CHAPTER 5: FINDINGS ................................................................................................................ 1255.1 An Overview of the Case study ............................................................................................... 1265.2 Framework 1: Analytical (E, P, I) ............................................................................................ 132

5.2.1 The Environment (E) .................................................................................................... 1335.2.2 The Person (P) .............................................................................................................. 1555.2.3 The Interaction (I) ......................................................................................................... 1635.2.4 Refinement of Framework 1 ......................................................................................... 179

5.3 Framework 2: Theoretical (Human Agency MTS) .................................................................. 1795.3.1 Meta-theory: Structuration Theory (ST) ....................................................................... 1805.3.2 Critical Theory: Social Theory of Gender (STG) ......................................................... 1845.3.3 Middle-range theory: Social Cognitive Theory (SCT) ................................................. 1855.3.4 Operationalised theory: Social Cognitive Career Theory (SCCT) ................................ 194

5.4 Framework 3: Ontological (Critical Realism) .......................................................................... 1945.4.1 Agent-driven mechanisms ............................................................................................ 195

5.5 Chapter Summary .................................................................................................................... 209

CHAPTER 6: DISCUSSION ........................................................................................................... 2116.1.1 What are the influences? ............................................................................................... 2156.1.2 Framework 1 (The guiding analytical Framework): E, P, and I .................................... 2176.1.3 Outcome: Sphere of Influence (SoI) Model .................................................................. 2346.1.4 Outcome: Events of Interaction (EOI) model ............................................................... 2366.1.5 Summary ....................................................................................................................... 2386.1.6 Framework 2: The Human Agency Multi-Theory Scaffold (HAMTS) ........................ 2396.1.7 Summary ....................................................................................................................... 2466.1.8 Framework 3: Ontological (Critical Realism) ............................................................... 2466.1.9 Summary ....................................................................................................................... 252

6.2 A synthesis of the three Frameworks ....................................................................................... 2526.2.1 The Complementary Value of the Three Frameworks .................................................. 2526.2.2 The focus on agent-driven mechanisms ........................................................................ 2566.2.3 Outcome: The Five Acts of Agency (FAA) Model ...................................................... 2606.2.4 Outcome: The Acts of Agency Theory ......................................................................... 279

6.3 Application of the Five Acts of Agency (FAA) Model ............................................................ 2866.3.1 Proposing Strategies by drawing on the Five Acts of Agency ...................................... 2866.3.2 Evaluating Strategies by drawing on the Five Acts of Agency ..................................... 2956.3.3 Summary ....................................................................................................................... 297

6.4 Chapter Summary .................................................................................................................... 300

CHAPTER 7: CONCLUSIONS ...................................................................................................... 3027.1 Addressing the research problem ............................................................................................. 304

7.1.1 The influences on women’s participation ..................................................................... 3057.1.2 Three Frameworks with which to Identify and Understand influences ........................ 306

7.2 Contribution ............................................................................................................................. 3177.3 Limitations of the research ....................................................................................................... 3197.4 Future research opportunities ................................................................................................... 3217.5 Final comment ......................................................................................................................... 323

Author: Anitza Geneve Year: 2013 P a g e viii of 401

List of IllustrationsFigure 1: Multimedia and games as a subset of core ICT industries (ACS, 2001; in Houghton, 2001) ................. 30 Figure 2: The DCI in relation to the ICT and traditional cultural industries .......................................................... 30 Figure 3: Games workers’ characteristics in Australia, June 2006-2007(Source: AIMIA, 2011-based on ABS data) ....................................................................................................................................................................... 35 Figure 4: Framework 1: Analytical Framework comprised of three categories (E, P and I) .................................. 55 Figure 5: The HAMTS theoretical framework (Framework 2) ............................................................................... 75 Figure 6: Framework 3: ontological framework comprised of three concepts from Bhaskar’s Critical Realist ontology ................................................................................................................................................................. 83 Figure 7: Illustration of the context and ‘unit of analysis’ of the reported case study ............................................ 95 Figure 8 Example of Nvivo project showing tree nodes ...................................................................................... 113 Figure 9: Refined ‘Environment’ (E) category .................................................................................................... 133 Figure 10: ‘Early career’ context within the social dimension of the ‘Environment’ (E) category ...................... 135 Figure 11: Macro and micro levels across the four dimensions of the Environment (E) ..................................... 137 Figure 12: Refined Person (P) category ............................................................................................................... 155 Figure 13: Components of an ‘Act of Agency’ .................................................................................................... 196 Figure 14: The Enabled ‘Act of Agency’ ............................................................................................................. 198 Figure 15: The Connected ‘Act of Agency’ ......................................................................................................... 199 Figure 16: The Collaborative ‘Act of Agency’ .................................................................................................... 200 Figure 17: The Creative ‘Act of Agency’ ............................................................................................................ 203 Figure 18: The Human ‘Act of Agency’ .............................................................................................................. 206 Figure 19: Visual representation of this (‘Discussion’) chapter ........................................................................... 214 Figure 20: Examples of the types of influences on women’s participation in the DCI ........................................ 216 Figure 21: Social settings and agents in the social dimension from a lifespan perspective .................................. 221 Figure 22: Extending the person category............................................................................................................ 227 Figure 23: Refinement of the first two categories in Framework 1 resulted in the Sphere of Influence (SoI) model ................................................................................................................................................................... 234 Figure 24: Positioning the analytical category of ‘women’ in the Sphere of Influence (SoI) ............................... 235 Figure 25: Development of the SoI and EoI models ............................................................................................ 237 Figure 26: Benefits of a Multi-Theory Scaffold (MTS) ........................................................................................ 242 Figure 27: The MTS supports Stage 3 of Danermark et al.’s (2002) model ......................................................... 250 Figure 28: The HAMTS assists in the abstraction of mechanisms identified in the Five Acts of Agency ............. 251 Figure 29: Extending Framework 1 to include the agent-driven mechanisms ..................................................... 257 Figure 30: Identifying agent-driven mechanisms by drawing on Critical Realism and agency theory ................ 258 Figure 31: Five Acts of Agency: Ten agent-driven mechanisms .......................................................................... 261 Figure 32: The elements and relationships involved in the Acts of Agency theory .............................................. 281 Figure 33: Model of Girls Career Choices (Adya and Kaiser, 2005). .................................................................. 355 Figure 34: Career stage model (Ahuja, 2002) ...................................................................................................... 355 Figure 35: Webb and Young’s (2005) major and sub-themes .............................................................................. 355

Author: Anitza Geneve Year: 2013 P a g e ix of 401

List of Tables Table 1 Summary of the three frameworks used to analyse the empirical data...................................................... 20 Table 2 Summary of research questions, aims, approach, and findings ................................................................. 22 Table 3 Overview of chapters ................................................................................................................................ 25 Table 4 Types of literature informing the research ................................................................................................ 28 Table 5 Game development workers by employment type and gender at end June 2007 (ABS, 2008a) ............... 34 Table 6 Participation rates of programmers in the Australian and international DCI ............................................ 36 Table 7 Theoretical approaches to explaining influences on women’s participation in ICT .................................. 48 Table 8 Examples of strategies/initiatives to foster women’s participation (a full table with further information is included in Appendix 3) ................................................................................................................. 49 Table 9 Conceptual framework offered by IDT (Trauth et al., 2004; Trauth, et al., 2005) .................................... 54 Table 10 The four types of theory in a Multi-Theory Scaffold (MTS) .................................................................... 67 Table 11 Framework 2: the HAMTScomprised of four Human Agency-related theories ...................................... 74 Table 12 SCT concepts (Bandura, 1997) evident in previous research .................................................................. 78 Table 13 Key Critical Realist authors informing the research approach ................................................................ 82 Table 14 Ontological assumptions of the Critical Realist view ............................................................................. 83 Table 15 The three frameworks to address methodological concerns .................................................................... 88 Table 16 Case study outcomes resulting from the research conducted in this dissertation .................................... 92 Table 17 Relevant characteristics of a case study approach .................................................................................. 93 Table 18 Context of the case study ....................................................................................................................... 94 Table 19 Examples of sampling strategies in the reported case study................................................................... 96 Table 20 Types of data in the case study based on Layder’s Adaptive Theory (1998) ........................................... 98 Table 21 Data source and collection method employed in the reported case study ............................................... 98 Table 22 Examples of secondary sources of data and reasons for their inclusion (Full list in Appendix 17) ...... 104 Table 23 Danermark et al.’s (2002) six stage model broadly aligned to the three frameworks............................ 106 Table 24 Summary of key analysis activities aligned to the three frameworks used, and to Danermark et al.’s (2002) model of explanation ................................................................................................................................ 114 Table 25 Generalisability of case study findings ................................................................................................. 118 Table 26 Summary of research design presented in Chapters 3 and 4 ................................................................. 124 Table 27 Case study participants (female DCI workers and industry stakeholders) ............................................ 127 Table 28 Overview of female DCI participants ................................................................................................... 128 Table 29 Overview of stakeholder participants .................................................................................................... 129 Table 30 Case study organisations (including games and multimedia production) ............................................. 129 Table 31 Lifespan stages and contexts of participation ........................................................................................ 134 Table 32 Matrix of influences across the four dimensions of the ‘Environment’ (E) category from a life stage perspective ........................................................................................................................................................... 136 Table 33 Examples of influence in the DCI within the environment dimension .................................................. 138 Table 34 Micro context related to lifespan stages ................................................................................................ 139 Table 35 Social agents identified within social settings ....................................................................................... 139 Table 36 Hours worked by Phase 1 participants .................................................................................................. 145 Table 37 Examples of the types of media and messages within the mediated dimension .................................... 152 Table 38 Examples of resources across a lifespan perspective ............................................................................ 154 Table 39 The 10 Events of Interaction manifesting in the ‘actual’ domain of reality .......................................... 164 Table 40 Female interactive content creators’ access to the DCI ......................................................................... 165 Table 41 Concepts from the Human Agency Multi-Theory Scaffold (HAMTS) .................................................. 180 Table 42 Empirical evidence of Giddens’ concept of modalities ......................................................................... 180 Table 43 Empirical evidence of Bandura’s person mechanisms .......................................................................... 186 Table 44 Empirical evidence of Bandura’s environmental mechanisms .............................................................. 191 Table 45 How the three frameworks illuminated the three ‘domains of reality’ .................................................. 195 Table 46 The Five Acts of Agency: The underlying agent-driven mechanisms .................................................... 195 Table 47 Five Acts of Agency: Empirical evidence of agent-driven mechanisms ................................................ 197 Table 48 Examples of the temporal aspect of influences on participation ........................................................... 235 Table 49 Linking the Events of Interaction to previous research ......................................................................... 238 Table 50 A summary of recommendations to foster the participation of women in the DCI ............................... 294 Table 51 The Five Acts of Agency and strategies for the event of motherhood .................................................... 295

Author: Anitza Geneve Year: 2013 P a g e x of 401

Table 52 Examples of agent-driven mechanisms (in the FAA model), linked to possible strategies ................... 296 Table 53 Use of the Five Acts of Agency model to evaluate Smart Women – Smart State Strategies Science, Engineering and Technology Action Plan, 2006 – 2009 (Queensland Government, 2006c) ............................... 298 Table 54 Examples of future areas of research resulting from this dissertation ................................................... 322 Table 55 Examples of strategies fostering girls and women’s participation ........................................................ 352 Table 56 Example of influences on participation as identified in ICT and DCI literature ................................... 353 Table 57 Teague’s (1997) application of the PRECEDE-PROCEED model ....................................................... 356 Table 58 Approach to establishing ‘trustworthiness’ of the research findings ..................................................... 357 Table 59 Example of key databases consulted for literature review .................................................................... 367 Table 60 Research design attributes identified in the meta analysis of key literature. ......................................... 367 Table 61 Types of interviews questions (based on Strauss et al., 1964, pp, 26 – 27) ........................................... 368 Table 62 Approaches to data representation implemented in the study ............................................................... 369 Table 63 Application of Giddens’ ST (Giddens, 1984, pp.281-28) and Shoib et al.’s (2006) recommendations towards applying Giddens ST .............................................................................................................................. 371 Table 64 Approximate participation rates for women working in the ICT, IT and DCI (Australia and internationally from 1997 to 2007). ..................................................................................................................... 401

List of Appendices Appendix 1: Publications stemming from the research ........................................................................................ 350 Appendix 2: Example of literature indicating women are under-represented in the DCI and related ICT industry in Australia and internationally .............................................................................................................. 351 Appendix 3: Examples of strategies fostering women participation .................................................................... 352 Appendix 4: Influences on women’s participation as identified in previous research ......................................... 353 Appendix 5: Models regarding women’s participation from the ICT literature ................................................... 355 Appendix 6: Approach to establishing ‘trustworthiness’ of findings ................................................................... 357 Appendix 7: Case study organisation vignettes ................................................................................................... 358 Appendix 8: Case study and interview protocols ................................................................................................. 359 Appendix 9: Online questionnaire ....................................................................................................................... 363 Appendix 10: Examples of interview question bank ............................................................................................ 364 Appendix 11: Hermeneutic principles applied in the research ............................................................................. 366 Appendix 12: Giddens’ Model of Interaction (based on Giddens, 1984) ............................................................. 366 Appendix 13: Search strategy details ................................................................................................................... 367 Appendix 14: Example of researcher reflection ................................................................................................... 368 Appendix 15: Methods of data representation used in the current study .............................................................. 369 Appendix 16: Reflection on the application of Giddens’ Structuration Theory (ST) ........................................... 371 Appendix 17: Secondary sources of data ............................................................................................................. 372 Appendix 18: Approximate participation rates for women working in the ICT, IT and DCI (Australia and internationally from 1997 to 2007). ..................................................................................................................... 401

Statement of Original AuthorshipThe work contained in this thesis has not been previously snbmitted to

meet requirements for an award at this or any other higher education institution. To

the best of my knowledge and belief, tile thesis contains no material previously

pnblished or written by another person except where due reference is made.

Candidate: Anitza A. Geneve

Signature:

Date: 24th May2013

’Women’s Participation in the Aostralian Digital Content Industry"Author: Anitza Geneve }’ear: 2013 P a g e .\’i of 401

Author: Anitza Geneve Year: 2013 P a g e xii of 401

Acknowledgments

“We create society at the same time as we are created by it” (Giddens, 1984, p.14).

“Society is only present in human action, but human action always expresses and utilizes some or other social form” (Bhaskar, 1998, p.36-7).

To my daughter Finn,

The little girl wandered up to her mother, who was sitting in front of a laptop working. With a beaming smile she exclaimed, “Mummy, when I grow up I’m going to make computer games”. Her mother looked at her, first with pride, but this quickly turned to a look of concern. What would this woman’s response to her daughter be? She wanted to say, “Darling, you can do anything you set your mind to”; instead the words, “Sweetheart that’s wonderful (not so believably) ...but along the way there might be things that make it hard,” tumbled out. After a thoughtful pause, the little girl asks, “Like ‘what’ sort of things mummy?” The mother thinks to herself: What are the obstacles? What first flashes through her mind are the media articles that regularly state that there are fewer women working in the games industry, that it is a male dominated culture, and that a woman in this industry is seen as an exception to the norm. She thinks of the emails she receives from various groups promoting strategies to encourage girls into ICT. She thinks of the academic debates of why there aren’t more women, which range from women just not being suited to technological work, to women being able to do it but being hampered by social expectations or (colloquially) ‘nature vs. nurture’. Slowly, she answers: “Well, sweetheart you will find that there are many things”. Following a moment of silence, the question that every little child asks comes: “But ‘why’ mummy?” The mother answered, “Well, that’s a very good question sweetheart”.

Parents may never know ‘everything’, but I can proudly say that I have contributed

to the world that my daughter will grow up in by asking some of the right questions. After

spending eight years of my life exploring this question, there is, and I suspect never will

be, one answer. Women’s participation in many facets of society remains a complex

phenomenon. I have been privileged to share the insights of those passionate women

working in the Digital Content Industry. I can say that these women do face obstacles to

their participation; however, their individual acts of agency can challenge and transform

such constraints. I hope the insights from my research can help industry and our broader

society to understand and support such personal agency.

I wish to thank my Principal Supervisor Professor Karen Nelson. Thank you for

your support in helping me reach a personal goal that has been so rewarding. I couldn’t

have asked for a better supervisor. It’s true: “That which does not kill us makes us

stronger” (Nietzsche). I wish to thank my Associate Supervisor Associate Professor Ruth

Christie for helping me start the journey. Dr Greg Timbrell’s and Dr Michael Middleton’s

involvement as Associate Supervisors was similarly invaluable. Greg’s enthusiasm

provided great intellectual energy when needed most and a most memorable quote

involving the word theorists’ and Black Sabbath in the same sentence.

Professionally, I would like to thank QUT for supporting my research in the form

of a QUTPRA scholarship, and for providing funding for various activities (such as

attendance at the Oxford Internet Summer School). QUT supported me not only as a

student, but also as a working mother. I would like to thank my fellow colleagues from the

Author: Anitza Geneve Year: 2013 P a g e xiii of 401

Oxford Internet Institute Summer School of 2009 for their positive feedback. Particular

thanks must go to those people in the multimedia and games industry who gave their time

and shared their insights. The involvement of the women working in the Digital Content

Industry and industry stakeholders has been integral to the research. Thanks must go to

Denise Scott who edited the final version of the thesis as per Australian Standards for

Editing Practice. Thanks must also go to my employer, Southbank Institute of Technology,

who have provided me with the flexibility to complete this study. And finally, thanks must,

of course, be offered to the colloquium panel and the final examiners for their collegial

insights.

Personally I would like to thank Jill for her editing, my friend Jenine who shared

insights from her own PhD journey and other friends for understanding the commitment I

have made to complete this research. Finally, thanks must go to my partner Simon, whose

ways of supporting me during my research would take half the space of the dissertation if I

were to list them all.

Chapter 1. Introduction ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 15 of 401

Chapter 1: Introduction

This dissertation, ‘Women’s Participation in the Australian Digital Content

Industry’, reports the investigation into the research problem: Why do women participate

in the Australian Digital Content Industry (DCI)? The need to address this research

problem arises from the fact that women are currently under-represented in technology-

related roles within the DCI workforce, both in Australia and internationally. There are

two key research questions the investigation addresses. The first question is: What are

the influences on women’s participation? The second question is: How can we

understand these influences? The research approach captures empirical data that provides

insights into the factors influencing women’s participation in the Australian DCI.

Over recent decades, academia, government, industry and the media have all

asked the question: Why are women under-represented in occupations associated with

technology? This question has most often surfaced within the Information and

Communication Technology (ICT) research (See the work of Trauth, Quesenberry &

Morgan, 2004; Trauth & Quesenberry, 2007; Trauth, 2011). However, despite a body of

significant academic research and discussion towards understanding the phenomenon of

participation, women continue to participate in inequitable numbers in ICT education and

career pathways. A sense of urgency permeates the wider literature, as industry journals

announce that “women are fleeing the technical professions” and “deserting IT studies

and jobs in droves” (Walters, 2006, p. 26). Hence, the “documented need to study the

gender imbalance” that Trauth (2002, p.98) recognised a decade ago, still remains.

There is a similar need to study the gender imbalance in the DCI, an industry of

emerging importance within the Australian economy over the last decade (Queensland

Government, 2003; Australian Government, 2004; AIMIA, 2005; DCITA, 2005, 2006).

The DCI is most readily identified with the production of multimedia interactive content

(such as websites, mobile content and augmented reality) and games production (for

example: Wii, Playstation and xBox, mobile games and serious games) 0F

1 As Section

2.1.1 explains, the DCI and the ICT industry share similarities. There are overlaps in

occupational roles and skills, such as programming, and both extensively draw on

technology. Like the ICT industry, the DCI is also characterised by the inequitable

participation of women. 1 In 2007, the Australian Bureau of Statistics (ABS) survey of ‘Work in Selected Culture and Leisure Activities’ separated games production and website development.

With little empirical or theoretical research of DCI workers’ experiences–or,

more specifically, of women's experiences– within the DCI, there is a need for continued

research efforts. Accordingly, this research aims to contribute to the body of knowledge

concerned with women’s participation in technology-related industries by investigating

the influences on women's participation within the Australian DCI. Influences may

include both those that constrain women’s participation, and those that foster

participation. Understanding influences on women’s participation within the Australian

DCI may contribute to addressing women’s inequitable employment in the industry.

This chapter introduces the researcher’s motivation (Section 1.1), the context of

the study (Section 1.2), the research problem and questions (Section 1.3), the research

approach (Section 1.4), the research contribution (Section 1.5) and an outline of the

remaining chapters (Section 1.6).

1.1 PERSONAL MOTIVATION

The impetus for this research gradually evolved from the researcher’s personal

observations as an educator and freelancer in the multimedia industry over the past two

decades. In the mid 1990s, there was an excitement about the technological, social,

economic and creative potential on offer in the newly emerging ‘new media industry’1F

However, following this excitement, a concern emerged that women were not entering or

remaining in the industry. In the early 2000s, the researcher was actively engaged in

encouraging women into vocational educational pathways, and was subsequently

recognised for these efforts2F

3 One initiative the researcher was involved in was the

Diploma of Multimedia for Women, which was identified by the Beyond the Pink

Collar: Towards Strategies to Respond to Women and Work Issues in Queensland

(Queensland Government, 2001) as a program addressing women’s inequitable

participation. However, the researcher held some concerns about the program. For

example, the females undertaking the program were charged approximately $500 more

than male students enrolled in an almost identical educational program; this extra cost

was associated with providing mentoring and role-models for the female students3F

By the mid 2000s, the researcher observed even fewer women entering the newly

emerging games educational programs. For example, in the first two years (2004-2005)

of the Diploma of Multimedia (Games Specialisation), not one student was female. The

number of female students enrolled in these programs is still low. For example, in July 2 The term ‘New media’ is taken broadly to mean the same as the term ‘Digital Content Industry’. 3 For example, the researcher was the recipient of a 2005 Unisys TAFE Equity Outcomes award. 4 These costs were associated with the extra support female students received, primarily in the form of guest speakers such as role model Dr Dale Spender. The researcher championed the parity of cost between the Diploma for Multimedia for Women and other non-gendered diplomas. The extra cost had been identified by the female students as adding to the financial burden of undertaking the program.

2011, only 2 of 16 multimedia students and 1 of 20 games students were females. These

observations and experiences motivated the researcher to understand the influences on

women’s participation.

Rather than investigating the participation of students within the education

context, the researcher focuses on women in their early careers within the industry

context. By focusing on the industry context, the research embraces Consalvo’s (2008)

approach of exploring the “future workplace for those girls who will be shaping

products” (p. 179). The researcher’s background as student, freelancer, teacher and role

model within the DCI is beneficial to her capacity to undertake this research. Benefits

include being able to closely identify with the participants’ experiences of the industry

context under study. When tempered by the (warranted) critical reflection regarding

potential researcher biases, this personal connection strengthens the researcher’s insights

into the phenomenon.

1.2 CONTEXT

It is important to define the context in which the phenomenon of women’s

participation manifests. Doing so helps define the boundaries within which findings have

application. In light of the research focusing on women’s participation in the DCI

workforce, it may seem at first that the pertinent contexts are DCI organisations

themselves. However, because the causes of women’s under-representation may be

found in the wider society (Trauth, 2006), there is a need to acknowledge the social and

historical characteristics of the broader context within which the DCI operates.

Delineation of the social context involves an acknowledgment of societal assumptions

surrounding women’s capacity for occupational roles, with a shift over the last few

decades from women inhabiting the ‘domestic sphere’ (Becker, 1985, Pocock, 1998, &

Alexandrowicz, 2005) to occupying non-traditional occupational roles in workplace

contexts. Delineation of the historical context requires “contextualizing how things came

to be” (Allen, 2009, p.3)4F

5, and a recognition of the salient characteristics of society. One

such characteristic is the presence and use of technology, which Negroponte (1999)

described as being a key hallmark of contemporary society. Technology warrants

specific recognition as the DCI and ICT industries are characterised by their adoption

and creation of such technology.

There is no shortage of rhetoric in the academic literature which recognises that

technology has been a transformational aspect in people’s lives, profoundly changing

“the manner in which we work, communicate and cooperate with each other...”(Säljö,

5 http://www.netcrit.net/presentations/oii-summer-doctoral-program-historicising-the-internet-presentation/

2004, p. 492). Indeed, the previous decade has provided us with significant

technological innovation and changing social practices. Social media applications such as

Twitter lead us to participate in society via online ‘twittering’, sites such as YouTube

form a new model of communication (Bruns, 2008), and Facebook has seemingly

become entrenched in everyday life in Australia.F

6 In short, much has changed this past

decade in regards to technology innovation and adoption.

The DCI and ICT industries are also industries in which women have historically

been, and currently remain, under-represented. In 1967, Seligsohn suggested that

Information Technology would be a gender-neutral field unencumbered with historic

gender stereotypes. He emphasised that the qualities of problem solving, attention to

detail, accuracy and patience were needed in computer programmers–qualities which

may even “give the girl programmer an advantage over her male colleagues” (Seligsohn,

1967, p.186). However, in 2003, women represented only approximately 20% of all

Queensland7F7 computing professionals (Queensland Government, 2003, p.10). Similarly,

research from the United States suggests that computing remains a heavily male-

dominated field, even after twenty-five years of extensive efforts to promote female

participation (McGrath-Cohoon & Aspray, 2006).

Across the global context, the DCI shares a similar story (as will be further

discussed in chapter 2). In 1998, Cassell and Jenkins (1998) in their canonical book,

From Barbie to Mortal Kombat, raised concerns regarding women’s participation in the

computer games industry. Then again in 2008, a decade later, there remained the need to

revisit the “debate about how more women could participate in the emerging digital

industries” (Jenkins & Cassell, 2008, p. 13). With Australian games development

organisations only employing 154 in contrast to 1,277 men, in the 12 month period prior

to June 2007, it appears there is indeed a need to ask why more women do not

participate. Both the nature of technology and its adoption practices have changed this

past decade; however, there is less change evident with respect to women’s equitable

participation within the ICT and DCI industries.

Thus the context of the reported research includes; a society where technology

and digital products have become more ubiquitous, a culture where women are more

likely to be in the workforce, but an emerging industry (DCI) where women are under-

represented in Australia. The research seeks to understand why women participate or do

not participate in the industry. This is an important question to ask as it has social and

6 YouTube http://www.youtube.com (began 2005), Twitter http://www.twitter.com (began 2006), Facebook http://www.facebook.com (began 2004). 7 Queensland is a state of Australia

ethical implications, as well as implications for the industry, where gender diversity may

improve innovation (as will be discussed in Section 2.3).

1.3 PROBLEM STATEMENT, RESEARCH QUESTIONS AND OBJECTIVES

The research problem is: Why do women participate in the Australian Digital

Content Industry (DCI)? As will be introduced in Chapters 2 and 3, two key research

questions can address the research problem.

RQ 1: What are the influences on women’s participation in the DCI?

RQ 2: How can the influences on women’s participation be understood?

As Chapter 2 discusses, the literature relevant to women’s participation in the

ICT industry suggests there is a plethora of influences on their participation. These

influences include characteristics of the environment (for example, structural) and of the

person (for example, psychological). There is, however, a paucity of either conceptual or

empirical research regarding the phenomenon of women’s participation in the Australian

DCI. This paucity of research and the need to identify those influences relevant to the

Australian DCI leads to the first research question: What are the influences on women’s

participation within the DCI? Research Question 1 may be answered with an analysis of

empirical data that illuminates the influences experienced by women employed in the

DCI. Therefore, the first aim of the research is for an empirically informed rich

description of the influences on women’s participation within the little-studied context of

the Australian DCI.

As Chapter 2 also discusses, influences must be understood in a unified manner,

particularly as previous research of women’s under-representation in the related ICT

industry has been critiqued by Quesenberry (2006) for offering a “wide range of

fragmented and difficult to holistically understand findings” (p.343). Concordantly, there

has been a call by leading researchers for a greater theoretical understanding of

influences. Trauth, Quesenberry and Huang (2009) emphasise that when aiming to

understand the complex phenomenon of women’s participation in ICT, the core

challenge lies in developing a theoretical understanding of the phenomenon; there is,

they claim, a “challenge of how to theorize gender under representation” (p. 477). Thus,

a second research question emerges: How can we understand the influences on women’s

participation?

As will be presented in detail in chapter 3, the research approach proposes

employing three different frameworks to identify and understand the influences on

women’s participation in the Australian DCI.

1. Framework 1–a broad analytical framework informed by existing research–

proposes three categories of influences: 1) the environment, 2) the person, and 3)

the interaction of the environment and person.

2. Framework 2–a theoretical framework informed by existing human agency

theories (as a scaffold)–is used to foster further understanding of the

environment, person and interaction.

3. Framework 3–an ontological framework informed by Critical Realism–focuses

attention on underlying mechanisms involved in interaction.

The frameworks, offering specific concepts for data analysis (as summarised in

Table 1), are used to sequentially analyse the empirical data collected as a part of the

exploratory case study. The findings reveal the frameworks ultimately foster a unified,

empirically informed, theoretical understanding of why women participate in the DCI.

Table 1 Summary of the three frameworks used to analyse the empirical data Framework 1. Analytical 2. Theoretical 3. Ontological

Basis of concepts in framework

Literature related to phenomenon under study

Four Human Agency theories in the Human Agency Multi-Theory Scaffold (HAMTS)

Critical Realism (specifically, Bhaskar’s notion of a stratified reality)

Specific concepts used in data analysis

Three broad concepts 1) Environment (E) 2) Person (P) 3) Interaction (I) of E and P

11 concepts from four agency theories.

Three domains of reality 1) Empirical 2) Actual 3) Real

In addition to the empirical and theoretical insights, findings may have applied

outcomes if they can be used to plan, support or evaluate strategies to foster women’s

participation in the DCI. As is raised in Chapter 3, although such strategies already exist

in the ICT context, the efficacy of interventions has been questioned (Craig, Fisher,

Forgasz & Lang, 2011). As Trauth et al. (2009) proposed, interventions may benefit from

theoretical insights. With this in mind, the research in this dissertation aims to inform

such strategies through theoretical insights built on empirical research.

1.4 RESEARCH APPROACH

The research design uses the established research approach of an exploratory case

study as a strategy of enquiry. Case study is congruent with the three frameworks, which

are used to investigate and understand the empirical data, as it can combine empirical,

theoretical and ontological insights to deliver descriptive and explanatory outcomes. The

scope of the case study is the experiences of women working in the Australian DCI. The

case study involves the collection and analysis of multiple sources of (mostly qualitative)

empirical data from 2007–2011. The primary data source is semi-structured interviews

with 18 female, early-career, interactive content creators employed in DCI organisations

in Brisbane, Australia. The Australian Bureau of Statistics (ABS, 2008a) defines the

occupational category of interactive content creators as those involved in ‘designing

websites and creating programs that contain film, sound, and animated components for

use on computers, electronic games and touch screens’. The case study is underpinned

by a Critical Realist philosophy (or ontological stance). This approach is innovative in

the sense that there are few examples of empirical Critical Realist case studies, and none

applied to research that investigates women’s participation in the DCI.

As Chapter 4 reports, both case study and Critical Realism have limitations for

data analysis and the Six Stage Model of Explanatory Research (Danermark, Ekstrom,

Jakobsen and Karlsson(2002) offers an approach that can strengthen the data analysis.

This model is highly congruent with empirical data, the use of existing theory and,

importantly, the ontology of Critical Realism. The actual data analysis techniques include

those more typical of case study, including qualitative approaches such as hermeneutic

analysis, which illuminate less visible influences. Recognition of less readily visible

influences is pertinent to research underpinned by Critical Realism, where the ultimate

aim is to abstract underlying mechanisms.

The case study is exploratory, in the sense that all three frameworks are guides

rather than fixed frameworks. The analytical framework provides three broad categories,

and the theoretical framework provides a ‘scaffold’. A scaffold is as Walsham (1995b)

and Layder (1998) identified, a guide based on existing theory and not an explicit

framework into which data must fit. The ontological framework directs attention to the

underlying causal mechanisms at play, which are not predefined. Even the ultimate aim

of a causal explanation is, from a Critical Realist perspective, open to fallibility.

Table 2 offers a summary of the key aspects of the research, including: the two

research questions that emerge from the review of existing literature (Chapter 2); the

three different frameworks guiding data analysis (as detailed in Chapter 3); and the

planned methodological approach (discussed in Chapter 4). Added in retrospect, the

summary also includes research findings (fully reported in Chapter 5) and the research

contributions (fully discussed in Chapter 6).

Table 2 Summary of research questions, aims, approach, and findings Research Problem

Why do women participate in the Australian Digital Content Industry (DCI)?Research Questions

1. What are the influences on women’s participation in the DCI? 2. How can we understand these influences?

Gaps to be addressed

Paucity of empirical/conceptual research within DCI context

Existing research suggests the need to focus on interaction

Research in related domain suggests gender under-representation is under-theorised

Need for insights into multi-level influences from a pluralistic paradigmatic approach

Research Aims

Gain empirical insights, primarily from female DCI workers’ perspective

Identify influences on women in the DCI experience

Move from description to explanation of the phenomenon by drawing on existing theory

Provide (via an ontological stance) explanation through integration of theory and empirical data

Methodological Approach: Three Frameworks

Analytical Framework comprised of three broad concepts: Environment(E), Person (P), and (I) Interaction

Theoretical Framework comprised of four agency theories, which form the Human Agency Multi-Theory Scaffold (HAMTS)

Ontological Framework comprised of concepts based on Bhaskar’s three ‘domains of reality’: Empirical, Actual, Real

Methodological Approach: Design and Methods

Exploratory case study; multiple sources of evidence

Build on existing research

Use of existing theories (more than one) to scaffold the research

Ontological stance draws attention to mechanisms underpinning interaction

Methodological Approach: Logic

Multiple logic from three frameworks

inductive abductive retroductive

Key Outcomes

Large volume of rich, descriptive empirical data; range of influences identified from the empirical data

Sphere of Influence and Events of Interaction Models guide understanding of empirical data

Multi-Theory Scaffold A typology of theories scaffold empirical research

Five Acts of Agency model offers a conceptual framework of underlying mechanisms and supports the development of the Actsof Agency theory

Key Findings

Range of influences Influences include environment (E) and person (P) characteristics; also include events manifesting from the interaction between E and P

Agency theory reveals influences such as a person’s self-efficacy; also assists in identifying mechanisms involved in the interaction of influences

Five categories of agent-driven mechanisms offer a way to understand women’s participation; the Five Acts of Agency model provides a way to apply or evaluate strategies

1.5 RESEARCH CONTRIBUTION

There are four primary types of contribution emerging from the research: 1)

empirical, 2) theoretical, 3) methodological, and 4) applied.

1) Empirical insights result from an exploratory case study collecting multiple

sources of empirical data. Industry stakeholders and women working in the DCI offer

rich descriptive insights of a little-researched context.

2) A theoretical contribution emerges from the development of several original

models resulting from the analysis of the empirical data and further conceptualisation

and theorisation.

3) A methodological contribution emerges in two ways. First, the research

provides an exemplar of an empirically grounded case study underpinned by a Critical

Realist philosophy. Second, the research proposes an original approach to the application

of multiple theories as a scaffold.

4) An applied contribution emerges from the application of findings to an

exploration of existing or potential strategies to address the issue of inequitable women’s

participation.

In summary, the research is of significant value in the following five ways.

1- It is significant to the research community because: a. it adds knowledge to a domain where there is little research

b. it draws on both empirical and theoretical insights to develop an emerging

theory and associated conceptual models

c. it augments the literature on the use of theoretical scaffolds

2- It is significant to women working in the DCI because: a. it provides women in the industry with a voice

b. it provides insights to support women working in the industry

3- The research is significant to the Digital Content Industry because: a. it identifies influences, which industries (that wish to encourage women’s

entry to the industry and to retain their participation) can take initiatives to

address

b. it focuses on fostering women’s participation which, in turn, can increase

workforce diversity and innovation

4- It is significant to those practitioners developing and implementing strategies to encourage participation of women in the DCI because:

a. it links empirical and theoretical insights to existing strategies

5- It is significant to the wider social domain because: a. insights can help to address social inclusion and social equity concerns.

1.6 CHAPTER SUMMARY

This chapter has introduced the research problem, which asks ‘Why do women

participate in the Australian DCI?’. Two research questions help to refine the research

problem. The two questions essentially ask, 1) what are the influences on women’s

participation and 2) how might we understand these influences and ultimately the

participation. These are important questions to ask as women are under-represented in

the DCI workforce, both in Australia and internationally.

The chapter has introduced the research design by: identifying that an exploratory

case study will be utilised; specifying the three frameworks that guide the data analysis,

and introducing the data analysis methods suitable for qualitative data. The case study

context is the Australian DCI, that is, organisations producing digital products and

content; however, context must also consider the wider social setting that the

phenomenon is situated within.

This chapter has presented the research in a fairly linear fashion. It may appear

that the research design, including the guiding frameworks for analysis, are applied in a

structured manner. However, the chapter does not truly indicate the complexity in the

overall research process, which has been more cyclical than linear. From the progression

of a simple research question of ‘what are the influences’ to the more challenging

question of ‘how can we understand these influences’, the process has involved an

openness to the data. From the progression of reviewing existing approaches to

developing original frameworks, the researcher has faced many analytical challenges.

From the progression of the analysis of data to the development of models and an

emerging theory, the process has required constant researcher reflection. The research

design will be further explained in chapter 3 and 4.

It is hoped that the research approach and findings help to further understand the

complex research problem ‘Why do women participate in the Australian DCI’. Chapter

1 has laid the research foundations by providing an overview of the research problem and

the approach taken to address the problem. Table 3 outlines the content of the following

chapters.

Table 3 Overview of chapters Chapter/ Key Aspects

Chapter Description

Chapter 1 Introduction

Establishes research problem

Chapter 1 introduces the proposed research. This includes identifying the research context, aims, questions and methodological approach.

Chapter 2 Women’s participation in the DCI

Reviews previous research

Chapter 2 presents the literature relevant to the research problem and the research issues. Providing an overview of women’s participation in the DCI positions the proposed research.

Chapter 3 Ways to understand influences Identifies research issues and approach

Chapter 3 presents an argument for the application of three frameworks to analyse empirical data. 1. Framework 1: a guiding analytical framework, emerging from existing research 2. Framework 2: a theoretical framework comprised of Human Agency theory

concepts. A Multi-Theory Scaffold (MTS) is offered in this chapter as an early contribution.

3. Framework 3: an ontological or philosophical framework comprised of Bhaskar’s three domains of a stratified reality: 1) empirical, 2) actual and 3) real.

Chapter 4 Research Approach

Articulates research methodology

Chapter 4 articulates the methodology with particular reference to the requirements of a case study underpinned by Critical Realism. It is proposed that Danermark et al.’s (2002) Six Stage Model of Explanatory Research can further assist data analysis. The chapter identifies how the research process addresses the research problem in a logical and reliable manner.

Chapter 5 Findings

Findings, Iterations of data analysis

Chapter 5 reports on three cycles of analysis, where each of the frameworks outlined in Chapter 3 is used to analyse the empirical data. Analysis leads to the proposal of three conceptual models: 1) The Sphere of Influence (SoI), 2) The Events of Interaction (EoI), and 3) the Five Acts of Agency (FAA).

Chapter 6 Discussion

Synthesis of insights

Chapter 6 discusses the findings and presents the descriptive, theoretical, methodological and explanatory contributions resulting from the research. The models emerging from the research as a contribution are discussed in reference to extant literature and the empirical data. These models include: the Sphere of Influence (SoI), Events of Interaction (EoI), Multi-Theory Scaffold (MTS) and subsequent Human Agency Multi-Theory Scaffold (HAMTS), and Five Acts of Agency (FAA). These models inform an emerging theory entitled Acts of Agency.

Chapter 7 Conclusions

Implications Chapter 7 presents the conclusions of the research in reference to the research problem and questions. The chapter also provides recommendations for future research.

Chapter 2. Women’s participation in the DCI ‘Women’s Participation in the Australian Digital Content Industry’ Author: Anitza Geneve Year: 2013 Page 27 of 401

Chapter 2: Women’s participation in the DCI

The previous chapter introduced the research. This chapter provides a review of

the substantive literature relevant to women’s participation in the Digital Content

Industry (DCI). Section 2.1 discusses the choice of literature used for the literature

review. Section 2.2 provides a background to the DCI, highlighting its historical

connection to the Information Communication Technology (ICT) Industry. Section 2.3

establishes that women are under-represented in the DCI and considers the impact of this

under-representation on society. Section 2.4 introduces the diverse range of influences on

women’s participation in both the DCI and ICT industries as identified in previous

studies. Section 2.5 raises a number of initial methodological concerns raised by the

literature with regard to a suitable approach through which to explore and understand

women’s participation. Accordingly, the chapter concludes with an argument for a

particular approach with which to investigate and understand the influences on women’s

participation. Chapter 3 then details this approach; an approach comprised of three

different frameworks to guide data analysis and further theoretical development.

2.1 THE CHOICE OF LITERATURE

The literature review encompasses previous research that reports on women’s

participation in DCI career pathways; it also encompasses research pertinent to women’s

participation in the ICT industry. There is a need for both of these research

concentrations for two key reasons. First, there is a strong historical connection between

the DCI and ICT industries (as noted in Section 2.2). It can therefore be assumed that

women in the DCI may face many of the same influences as women in the ICT industry.

Second, although there has been a growing body of international research regarding the

DCI, there is little research regarding the Australian DCI. There is even less research

regarding the under-representation of women with regard to their pathways to

employment, their career progression, or their retention in the industry (as will be

discussed in full in Section 2.4). In contrast, there have been, and continue to be, a

number of collections of high-quality empirical studies related to under-representation of

females in the ICT sector, and in the wider context of the Science, Engineering and

Technology (SET) industries. Indeed, Trauth et al. (2009) describe the research

concerned with women’s participation in the ICT industry as a “volume” (p. 477). Thus,

the literature from the ICT field can address the paucity of literature relevant to the DCI.

Trauth et al. (2009) emphasise that the under-representation of women in

industries associated with ICT presents researchers with a complex phenomenon.

McGrath-Cohoon and Aspray (2006) recognise that the issue of women’s equitable

participation in Information Technology (IT) is so “complex” that it makes “it difficult to

know how to go about reaching a gender balance” (p.viii). Trauth et al. (2009)

emphasise that the under-representation of women in industries associated with ICT

presents researchers with not only a complex phenomenon, but one that “requires equally

complex solutions” (p. 477). Contributing to this complexity is the cross-disciplinary

nature of previous research.

Reflecting this complexity, the research approach involves the review of a multi-

disciplinary range of literature related to the topic of women’s participation. As Table 4

indicates, this literature includes both academic and non-academic sources. Academic

literature includes gender-related research (broadly defined as ‘feminist research’) and

ICT socio-technical literature. This academic literature often arises from different

schools of thought, and even from competing paradigms. Non-academic literature

includes government and industry reports, such as data from the Australian Bureau of

Statistics (ABS).

Table 4 Types of literature informing the research

Type Purpose Description

Academic literature

Domain-related and topic-related

Conceptual research

Studies that formulate emerging concepts, models and frameworks

Empirical research Studies with empirical data collection and analysis

Non-domain or topic-related

Wider theoretical frameworks

Theoretical approaches from literature outside of domain; for example, sociology

Non-academic literature

Public domain Blogs, government reports

Industry-based Reports or research from industry, industry websites

Furthermore, women’s participation is a sociological concern; thus, the literature

extends to domains such as sociology and psychology. A multi-disciplinary approach to

the choice of literature is vital for research which aims to understand a complex social

phenomenon, such as women’s participation, particularly if the aim is to offer insights of

value to both the academic community and to industry. As Trauth (2011) reminded us, a

research approach encompassing “boundaryless literature searching” is a strength (p.6).

2.2 AN OVERVIEW OF THE AUSTRALIAN DCI

This section provides an overview of the Australian Digital Content Industry

(DCI). The term ‘DCI’ does not often appear in the literature; however, it has been used

extensively by the Australian Government over the previous decade. Key government

reports include the Australian Digital Content Industry Futures (DCITA, 2005). Other

terms appearing in the literature include ‘digital interactive entertainment’ industry

(Martin & Deuze, 2009)8F

8 and ‘new media’ (Perrons, 2003a), although the latter term is

seen as being contentious (Pratt, 2000).9 The DCI is typified by the production of digital

cultural artefacts such as websites, mobile content and computer games, with key areas

of growth being in the development of games for console formats such as Playstation,

Xbox, and Wii (ABS, 2008a). Essentially, the DCI produces digital products and content

that are interactive and, in general, utilises technology in the process.

The growth of the Australian DCI continues to be a primary objective for both

the Australian Government and the industry itself. In 2005, the Australian Government’s

‘Digital Industry Content Action Agenda’ (Unlocking the Potential, DCITA, 2005)

estimated the number of employees within Australia’s digital content industry at

approximately 300 000. It also reported that the DCI contributed $A21 billion–almost 3.5

per cent of Gross Domestic Product–to the Australian economy. The agenda anticipated

the opportunity for the industry to double in value by 2015. In parallel, Australian

society has moved towards the ubiquitous use of digital content as reflected by the

uptake of mobile devices, such as smart phones and tablets (Nielsen Online Consumer

Report, 2010).10

There is a strong historical connection between the DCI and the ICT Industry in

Australia with considerable overlap across the products produced, the skills required by

the industry, and workplace practices. For example, in 2001, a report from the Australian

Computer Society (ACS) positioned multimedia and games as a component of the core

ICT Industry (See Figure 1). Consequently, much early industry and academic research

regarding the DCI has appeared under the ICT umbrella. This is also true for research

concerned with women’s participation in the ICT industry. For example, a Queensland-

based survey of women in ICT included a number of questions regarding respondents’

experiences with digital media or multimedia (Anderson, Lankshear & Klein, 2006).

8 For example, universities offering degree level programs in Games and Interactive Entertainment include Queensland University of Technology. 9 These terms are used interchangeably in this dissertation. 10 http://www.ad-tech.com/sydney_uploads/nielsen_WhitePaper2011.pdf

Thus, the literature regarding women’s participation in the ICT industry is relevant to

any study concerned with women’s participation in the DCI.

Figure 1: Multimedia and games as a subset of core ICT industries (ACS, 2001; in Houghton, 2001)

The DCI is also associated with the traditional cultural industries. Those working

in these cultural industries (Flew & Cunningham, 2010; UNCTAD, 2008) are exemplars

of Florida’s (2002) ‘creative class’ (Cunningham, Cutler, Ryan, Hearn & Keane, 2003),

and form the creative potential of Australia’s workforce (Creative Industries, a Strategy

for 21st Century Australia, 2011). 9 Hence, as Figure 2 illustrates, the DCI can be

conceptualised as spanning both the ICT and traditional cultural industries. The DCI has

“the applications and services components of the ICT industry on the one side and the

traditional film, entertainment and cultural industries on the other, and overlapping key

areas of both” (DCITA, 2006, p. 8). Consequently, DCI workers require both creative

talent and software specific skills (AIMIA, 2005. p. 41; Roan & Whitehouse, 2007).

Figure 2: The DCI in relation to the ICT and traditional cultural industries

Occupational roles typically found in the DCI include: web developer, games

designer and programmer; less typically roles include audio developer, social media

coordinator and graphic designer. There are few guidelines regarding salaries, with there

being no specific industrial relations awards. This often leaves for variance in salaries

between similar roles and across organisations. However, resources such as the

esalarysurvey website provide some insight, indicating that the base salary of a junior

web developer is $49,000 (AUD). Recently, a peak industry body, the Digital Industry

Association of Australia (AIMIA), conducted salary surveys. 11 DCI organisations can

range from large multinationals through to a start-up comprised of one to two workers.

Occupational roles in the DCI are difficult to define. In part, this is because

roles overlap with other industries (as noted), furthermore many DCI occupational roles

have only recently emerged. It was only in 2001, that the Australian Culture and Leisure

Classifications (ACLC) defined the key occupational role within the DCI as that of

‘Interactive Content Creator’ (Class 267; Cat. No. 4902.0), which includes those people

employed as web designers and games developers (ABS, 2008b; see Appendix 7 for

ABS definition). Even the ACLC presents a somewhat out of date listing of the types of

digital products an ‘Interactive Content Creator’ may develop; for example, there was no

mention of games for mobile phones in 2009. Roles such as ‘social media coordinator’

have only appeared with the advent of social media over the last five years or so.12 The

DCI occupational roles appear to be continuing to emerge.

There is a need for a clearer understanding of the DCI workforce both

internationally and in Australia. Although data regarding the participation of Australian

DCI workers is for the most part quantitative, there are difficulties in ascertaining precise

participation figures for those working in creative industries (Higgs & Cunningham,

2008; Cunningham et al., 2003). Issues regarding quantifying employment numbers are

also raised in the United Kingdom (Bakhshi, Freeman & Higgs, 2012; Roodhouse, 2006;

Pratt, 2000) and New Zealand (Walton & Duncan, 2002). In addition, quantitative data

does not provide insight into the finer nuances of the phenomenon of participation. For

example, Table 5 indicates that the first ABS survey of Australian games organisations

identified that most workers are employed fulltime. However, these figures do not

indicate that full-time employment may only last the life of a specific project production

cycle. Deuze, Martin & Allen (2007), for example, noted that employment in games

11 http://www.esalarysurvey.com and http://www.aimia.com.au 12 The researcher interviewed a previous student, who now works in the role of social media coordinator. This role did not exist at the time the interviewee was a student.

production is often contingent on the limited life of particular projects. In reference to

new media production, Gill (2002) noted that: “lives are organized around projects not

careers” (p. 73). Additionally, the extent of the use of contract multimedia workers has

differed across work areas such as creative, technical and educational areas (Swanson &

Wise, 1996).

All of the above factors–the importance of the Australian DCI, the emerging

nature of the DCI, and the limitations of the data about the DCI workforce–suggest that

there is a need for its further investigation.

2.3 CURRENT PARTICIPATION OF WOMEN IN THE DCI

The previous section provided an overview of the Australian DCI industry. This

section now provides evidence of the phenomenon of women’s under-representation in

multimedia and games production organisations in the Australian DCI. This evidence

strengthens the argument for further research into women’s participation.

2.3.1 WHAT IS PARTICIPATION

Broadly, ‘participation’ is defined differently across different domains.

Participation has been researched in contexts such as adult learning (Cookson, 1986;

Scanlan & Darkenwald, 1984), politics (Verba & Nie, 1972), and social inclusion, where

inclusion occurs through participation (Saunders, 2002). Theories such as ISSTAL

(Smith, 1980; Cookson, 1986; Brown, Grollmann & Tutschner, 2004) aim for an inter-

disciplinary approach. Another approach to understanding participation lies in labour

workforce participation data, where participation is “the extent to which the population is

willing and able to work” (Australian Government, 2005a) and participation rates refer to

the labour force expressed as a percentage of the population (Australian Government,

2005b). As Table 5 reveals, women’s participation in the DCI may be understood in

reference to participation rates in the workforce. Participation can also refer to under-

utilised labour (Abhayaratna & Lattimore, 2006).

There are however limitations in understanding participation only in terms of

participation rates. Rates themselves do not explain the phenomenon of women’s under-

representation in the DCI workforce. Research of women’s participation in the ICT

industry proposes that there is a need to look “beyond statistics” (Beekhuyzen &

Clayton, 2004, p.4) and “uncover the story beneath” (Trauth & Quesenberry, 2005). This

is particularly necessary as “little is known about the challenges these women experience

in the workplace in relation to their career stage” (Warne, Bandias, & Fuller, 2011). The

implication for this investigation is that there is a need to collect empirical data which

captures rich insights from women working in the DCI, and which can illuminate the

story behind the phenomenon.

2.3.2 WOMEN’S INEQUITABLE PARTICIPATION

Women’s inequitable participation is not a new phenomenon, having been cited

as a concern in the literature of different disciplines, including ICT and gender studies,

for a number of decades. Women have historically been under-represented as both users

and developers of digital products.

Research in the 1980s and 1990s focused on women’s participation as users.

Access to technology and digital products such as the internet (Turkle, 1995; Spender,

1995) was a central concern. Even those who heralded interactive computer networks,

such as Castells (1996) recognised the digital divide where certain groups, including

women, faced inequitable access to these technologies.

This early research also reported the constraints women faced. Cassell and

Jenkins (1998) reporting that women were seen as not being users of digital products –

‘girls did not play games!’ In contrast, recent research indicated a growing parity

between genders in the use of digital products such as websites (Women on the Web

Report, 2010). With regards to computer and console games, the Interactive Australia

2009 Report (Brand, Borchard, & Holmes, 2009) identified “that gamers would be

equally female and male by 2010” (p. 16). It appears that women, along with society,

have embraced the use of digital products such as games and websites.

A concern remains, however, that women are limited to being consumers rather

than creators of technology (Witkowski, 2006). Although the constraints that women

face accessing ‘cyberspace’ have largely disappeared, what remains is a concern that

women are still vastly under-represented in the fields that design digital technology

(Jenkins & Cassell, 2008). As a female Australian games developer noted, there is “still

little female input in games development” (Appendix 17, Item 49). Accordingly, there is

a need to consider women’s involvement as creators, and not only as users of digital

content.

There is ample evidence suggesting that women’s participation in the DCI and

the ICT industry is not equitable in comparison to men’s. Appendix 2 presents previous

research that has identified that women are under-represented in ICT careers, and in

educational pathways to those careers. Appendix 18, in summarising several sources of

information regarding women’s participation rates across industries related to

technology, suggests that their participation rates range from 5% to 30% (dependant on

the industry sector), both in Australia and internationally. For example, the first study by

the International Game Developers Association (IGDA), which surveyed nearly 6500

international games developers, identified women’s participation rates of approximately

11.5 % (Gourdin, 2005). A 2006 UK study identified a similar women’s participation

rate of 12% in computer games development (Pratt, Gill, & Spelthann, 2007).

The Australian Work in Selected Culture and Leisure Activities Report (ABS,

2007) identified that there were 62 600 males and 12 800 females in computer games and

other interactive software production–a participation rate of approximately 17%.13 The

inequitable participation rates are most apparent in games organisations. In games

development, “male workers almost completely dominate the core content creation roles

(such as design, programming and visual arts)” (Deuze et al., 2007, p. 346). The ABS

special report Australian Games Development Industry, in capturing workforce data for

the 2006-07 financial year (ABS, 2008a), showed that of the 1431 persons employed

Australia-wide in games development, only 154 were women (Table 5). Therefore,

women comprised 10.7% of all games workers in the Australian DCI during that

period.14 As Figure 3 indicates, the State of Queensland employed the greatest number of

games workers in Australia in that period. These rates clearly indicate that women are

under-represented in the Australian DCI workforce.

Table 5 Game development workers by employment type and gender at end June 2007 (ABS, 2008a)

Male no. % Female % Person %

Working proprietors/ partners/ salaried directors 45 3.5 3 1.9 48 3.4

Other employees

Permanent full-time 1,188 93.0 137 89.0 1,325 92.6

Permanent part-time 15 1.2 7 4.5 22 1.5

Casuals 29 2.3 7 4.5 36 2.5

Total 1,232 96.5 151 98.1 1,383 96.6

Employment at end June 1,277 100.0 154 100.0 1,431 100.0

Women’s participation in multimedia production roles appears to be higher than

for games production. In the UK, a survey identified a 38% women’s participation rate in

web design and special effects. In Australia, the Work in Selected Culture and Leisure

13 The report notes that the estimate has a relative standard error of 25% to 50% and should be used with caution (ABS, 2007). 14 It should be noted that there is a discrepancy in the ABS data as the total male workers added to female workers is 1386 and not 1431. Therefore, the suggested participation rate is an approximation.

Activities Report (ABS, 2007) identified 165 000 males and 89 200 females involved in

website production–an approximately 35% participation rate. The figures given above,

therefore, indicate that women’s participation rates may vary across DCI organisations.

The implication for this investigation is that a sampling strategy that captures data from

women working in both games and multimedia organisations must be employed.

Figure 3: Games workers’ characteristics in Australia, June 2006-2007(Source: AIMIA, 2011-based on ABS data)

Previous research also indicated that women face horizontal and vertical

occupational segregation. Although there were “ relatively high numbers of women in

‘web/multimedia’ jobs” in the Australian DCI, many of these appeared to be “primarily

clerical jobs”, and “women were rarely found in the more directly technical roles of

computer games design” (Roan & Whitehouse, 2007, p. 31). A similar segregation has

been identified in the Australian IT industry (Whitehouse & Diamond, 2005), the UK IT

industry (Panteli, 2005) and the UK games industry (Prescott & Bogg, 2011).

International industry-based research, such as the American Game Career Guide (2007),

identified that women are under-represented in developmental roles, including

programming, and are concentrated in high paying production roles. Thus, women’s

participation within the DCI may vary across occupational roles. The implication is that

this investigation needs to employ a sampling strategy that recruits women working in a

variety of roles within the interactive content creator category.

Industry-based reports indicate that women are particularly under-represented in

technical roles. Although there is no definitive definition of what a technical role

encompasses, it appears to include the skill set of coding or programming. As Table 6

indicates, ABS figures suggest programming is one of the key roles in games

development, with 29.1% of all workers employed as programmers. However, the ABS

figures do not indicate the number of women employed in programming roles. As Table

6 shows, an international survey of games production workers suggested that women’s

participation as programmers may be as low as 5% (Gourdin, 2005). A study of women

working in the UK games industry identified that women comprised only 2% of

programmers, 3% of audio, 5% of games designers, 8% of production staff, and 9% of

artists (Haines, 2004). Thus, women’s inequitable participation appears to be most

evident in roles related to technical skills. The implication is that this investigation needs

to employ a sampling strategy that recruits women employed in technical roles in the

Table 6 Participation rates of programmers in the Australian and international DCI

Left: Participation rates of both men and women in DCI (games) roles in Australia 2006-2007(ABS, 2008a) Right: International participation rates for games workers by gender and role type (Gourdin, 2005)

2.3.3 WHY PARTICIPATION MATTERS

There are many important reasons why there is a need to understand women’s

participation in the DCI workforce. In light of women’s under-representation, there is the

reason of social equity, which assumes each individual is given a fair and equitable

chance to pursue their passion. Such concerns of justice and equal opportunity has been

raised in the ICT industry (See Barker & Aspray, 2006). Having more women in the

industry can also address skills shortages. Indeed, initiatives aiming to foster women’s

participation in the ICT industry often suggest that greater participation by women may

help to address the skill-shortages the ICT industry faces (See, for example, the

partICipaTion National ICT Skills Summit issues paper, Queensland Government,

2006a). The 60Sox Report. Vol. 1 (Haukka, 2009) identified that the Australian DCI

faces a similar skills shortage, particularly in highly skilled workers. The report suggests

that the “Australian digital content industries often face skills and labour shortages,

exacerbated by insufficient supply of high quality industry-ready graduates”(Haukka,

2009, p. 5). However, there has be no call to recruit women as a way to address the skills

shortage in the Australian DCI.

These reasons do move beyond arguing simply for equitable workforce numbers

to acknowledging the value of having women in the DCI workforce. There is the

argument that a diversified workforce leads to greater innovation in both workplace

practices and in the types of digital products that are produced. Similarly, the ICT

literature proposes that fostering equitable participation of women (and other

marginalised groups) may improve workforce diversity and design innovation, including

increasing creativity and productivity (Bulkeley, 1995) and usability of products

(Sneiderman, 2000). Historically, the games industry had no interest in “thinking

seriously about gender” (Jenkins & Cassell, 2008, p. 10) even though a diverse

workforce is seen as “important” to its “future success” (Gourdin, 2005, pp. 9–10; Pratt,

2007)26F15. These arguments extend to the inclusion of women due to their gender-

specific skills, for example their soft skills. However, there is a need to recognise that

these gendered skills do not acknowledge the individual differences women possess.

These individual differences must be recognised without ascribing them to ‘nature’ (see

for example, Trauth, 2002).

There is also the reason that women’s participation may influence the nature of

the products produced, although this has been challenged by researchers such as

Consalvo (2008). Women’s under-representation as users of digital products, such as

games, has been attributed in part to the lack of appeal or suitability of the digital

products (Taylor, 2008; Beasley and Standley, 2002; Pham, 2008). 24F

16 Fullerton et al.

(2008) suggest that more women entering the industry may foster the development of

games that encourage more women players and, consequently, encourage more women

into the industry, thus forming the “virtuous cycle” (p.141). That there are fewer women

in the DCI workforce may well be influencing the types of digital content produced and

this in turn may dissuade girls and women from pursuing careers in the industry. This is a

similar argument to the one Wilson (2004) raises for the development of products in the

Information Services (IS) sector, where the “technology itself comes to be gendered

15 Interestingly, although Pratt is an academic, this quote appears in an online magazine (Computerworld) article. 16 Games such as Grand Theft Auto have received considerable media coverage for their violent and negative portrayal of women.

through the process of its design, development and diffusion into organisations and

society as a whole” (p.81).

Be it a combination of these reasons, or even other reasons yet to be considered,

the current research suggests that there is a need to better understand the phenomenon of

women’s participation, particularly at a time when society embraces technology and

digital products. Heeter, Egidio, Mishra and Winn and Winn (2009) suggest there may

well be a “high social cost” of the “female population not being engaged early in this

digital revolution” (p.97).

2.4 INFLUENCES ON WOMEN’S PARTICIPATION

Section 2.3 established that women are under-represented across interactive

content creator roles in multimedia and games production organisations within the

Australian DCI, particularly in technical roles. The section also identified several reasons

why it is important to address the under-representation. This section (2.4) reviews

existing literature to provide insights into what the influences on women’s participation

A review of a range of literature relevant to the phenomenon of women’s

participation in the DCI and ICT identifies a wide range of influences on women’s

participation (Appendix 13 provides details of the review procedure.) This review of

selected literature served the primary purpose of helping sensitise the researcher to the

range of influences that could be expected to surface in the investigation. In addition, the

review sensitised the researcher to methodological considerations and research gaps;

these are introduced in Section 2.5.

In brief, the review involved searching journal databases for peer reviewed

journal papers, using key search terms such as ‘gender, games’.17 Subsequently, snowball

sampling from the citations in key papers used techniques such as backward and forward

searches (Levy & Ellis, 2006) to identify research reported at conferences and in lower

ranked journals. Characteristics of each paper were captured in a Nvivo18 project and

included the following nodes: 1) influences explicitly identified, 2) empirical or

conceptual research, 3) the industry field (for example, ICT, DCI), and 4) the cohort (if

an empirical study).

17 For example, a database such as Ulrich’s indicates if a journal is ‘refereed’ or ‘peer reviewed’ http://ulrichsweb.serialssolutions.com/ 18 As is discussed further in chapter 4, Nvivo is a software tool used in data management and analysis

The review process revealed that there are few empirical academic studies

specifically investigating women’s participation in the DCI. Internationally, the notable

exceptions include Gill (2002) who focused on new media; Prescott and Bogg (2010)

who focused on games development; Perrons (2007) and Perrons, McDowell, Fagan,

Ray and Ward (2007) who focused on new media; and Krotoski (2004) who focused on

games. There are only two key Australia-based research papers relevant to women’s

participation in the DCI. Swanson and Wise’s (1996) findings (from a national survey of

the patterns of training, skills, income and employment of women working in the

multimedia industries) noted that although the industry area was “still” a male-dominated

industry, “women stand to gain from their participation in this expanding sector” (p. 3).

A decade later, Roan and Whitehouse’s (2007) study of DCI organisations in

Queensland, Australia, identified that women remain under-represented in multimedia

and in the emerging areas of game production. Gill (2002) has suggested, new media

workers are “rarely studied” (p.75); it appears, a decade later, that research into female’s

employment in these areas is even rarer. 19

Although there is limited research within the DCI field, a range of influences on

workers participation can been identified. These influences include the characteristics of

the environment and the characteristics of the person. Environment-based characteristics

include working conditions, workplace culture, and industry practices. For example, in

the European context, Gill (2002) identified that the environment characteristics

influencing new media workers include the changing patterns of work. Gill (2002)

recognised that lifelong employment is no longer the norm; this is, perhaps, best

illustrated by Flores and Gray (2000), who stated that the lives of “wired people” are

more like collections of short stories than the narrative of a bourgeois novel (pp.23-4).

The industry is characterised by periods of intense employment and long hours, followed

by quiet periods or periods of unemployment; indeed, Pratt (2000) referred to a career

with this pattern as a ‘bulimic career’. The environment characteristics that positively

encourage participation of both men and women include: the ability to work

autonomously with no managerial control, flexible working hours, and intrinsically

challenging and fulfilling work (Gill, 2002). However, the same characteristics could add

“an individualisation of risk” (Gill, 2002) and this may be a reason why workers leave

industry.

19 In this dissertation, the terms ‘DCI worker’ and ‘new media worker’ are interchangeable.

Some influences are more pertinent to women. Recent research regarding

women’s participation in games development organisations in the United Kingdom

identified a range of environment characteristics as constraining women’s participation.

Bogg and Prescott (2011) proposed that several characteristics of the games industry

could lead to an industry typified by young, male, unattached workers. These

characteristics include: women’s limited networking opportunities and access to mentors

and role models, limited flexibility and childcare provision, gender

discrimination, gender segregation of the workforce, higher values placed on masculine

attributes, a culture of long hours, and the potential need for workers to relocate.

Certain characteristics that may initially appear relevant to either gender require

further consideration regarding their influence on women in specific. Gill (2002)

recognised that a characteristic such as informality of work practices and relationships

could be both a positive characteristic that appeals to workers, and one that can also be a

negative characteristic if it contributes to gender inequity. Informality, for example, can

lead to environments where women experience inappropriate behaviour from male

colleagues. This informality also means that work is often allocated through

interpersonal connections, and the male majority could more readily foster these

connections. Thus, an influence can be both constraining and supportive.

Resources in the environment, such as digital games, may be particularly

pertinent when considering women’s participation. For example, having access to

resources that foster skill development and an individuals identification with the industry

is an important influence for women (Adya & Kasier, 2005). However, resources such as

game products may not appeal to women (Klawe, 2002) and certain areas of the DCI,

such as games development as seen as creating content for men by men.

The literature indicates that the environment-based influences women face in the

ICT and DCI manifest in different contexts, ranging from occupational or workplace

contexts to the wider social context. Influences that constrain participation can include

culture, which encompasses: general attitudes or values regarding women (Trauth,

Quesenberry, & Yeo, 2005); a woman’s cultural background (Trauth, Quesenberry, &

Huang, 2008); and the organisational culture in which a woman finds herself (Trauth,

Quesenberry, & Huang, 2009; Webb & Young, 2005). Organisational culture

characteristics, such as the dominant male culture in IT organisations (Webb & Young,

2005 citing von Hellens, Nielsen, & Trauth, 2001) and masculinised workplaces

(Griffiths, Moore, & Richardson, 2007), can leave women feeling as though they are in

the gendered minority. Beyond organisational culture lie the societal cultural norms,

which include gender and occupational stereotypes that suggest that IT workers are male

‘nerds’ and ‘geeks’, and that these stereotypes ‘turn women off’ seeking IT jobs

(Stockdale & Stoney, 2008; MacKnight, 2001). The media play a role in perpetuating

these inaccurate gender and occupational stereotypes (Jepson & Perl, 2002). This is so

even if the stereotypes are inaccurate, as suggested by Courtney, Lankshear, Anderson

and Timms (2009).

In addition to the environmental characteristics, there are also person-based

characteristics that influence women’s participation. Although there is less research

suggesting that the individual plays a role in their participation, it is evident that

individual characteristics are an influence. With reference to factors such as personality,

Bogg and Prescott (2012)–in citing Swanson and Woitke, 1997– recognised that barriers

to participation can be as much personal, as they are environmental. Personality

characteristics such as being strong are seen as important for women working in the IT

industry (Trauth at al., 2005; Trauth, Quesenberry, & Yeo, 2008); being ‘strong’ may

lead to perseverance in the face of environmental barriers. Cognitive qualities of the

individual that may influence a woman’s participation include: personality traits (Trauth,

2002), their actual skill capacity or gaps in experience (Margolis & Fisher, 2002), and

their perception of skills capacity or self-efficacy (Hackett, 1995). Motivated individuals

with a high sense of self-efficacy could well be more driven to participate than those with

low skills or low belief in their skills.

These individual characteristics, such as personality, motivation and self-

efficacy, often interact with the environmental influences. There appears to be a complex

interaction between influences.For example, environmental characteristics such as

financial reward or social benefits may motivate individuals. Yet, McGrath-Cohoon and

Aspray (2006) found that working in the IT industry was not appealing to women: it was

perceived as not offering any social value to society and women wanted careers that

made a difference. With regard to motivating workers through extrinsic rewards such as

financial reward, the ICT industry offers little incentive, with industry employment often

contract-based or volatile (consider, for example, the dot.com crash), which adds to

employee insecurity. The implication for this investigation is the need for a research

approach that can recognise both environment and person-based influences.

Gill (2010) recognised the role that individuals play in managing their own

careers, and described a new media worker’s life as a ‘pitch’–a theme that has appeared

in her earlier (2002) work. Gill also recognised that women face a challenge in

reconciling individual discourses that conflict with the sociological patterns of inequity.

That is, although women may recognise gender-based inequity, they rationalise their

experiences as individual, rather than direct consequences of environmental

characteristics. When individualistic understandings dominate sociological ones in this

way, disappointing (and even discriminatory) experiences are interpreted as personal

failures, or as random events (Gill, 2002). Gill recommended further research to consider

why women working in the new media industry do not embrace the gender discourse.

Gill’s work emphasised two salient points: first that the individual plays a role; and,

second, that women in new media do not readily assimilate their experiences to a

‘gendered’ experience. Gender is a hidden characteristic. As a respondent in a survey

conducted by Gill (2002) stated: 'You don't talk about gender if you want to get on'. Gill

(2002) suggested that these different perspectives or this evidence leaves the researcher

facing a ‘tension’ when exploring women’s inequitable participation. The implication

for this investigation is a research approach that is sensitive to gender.

Influences manifest in complex combinations. In the ICT industry-related

research, Trauth (2002) suggested it is a combination of individual, institutional and

cultural influences that result in women’s under-representation in the ICT industry.

Similarly, women working in the UK IT industry have identified “a combination of

factors”, rather than one single reason, as contributing to their leaving the IT industry

(DTI, 2005, p.16). An assumption of this investigation is that women in the DCI face a

similar range of influences on their participation, and that these environment-based and

person-based influences manifest in complex combinations.

This initial review process revealed a number of implications for the research. It

is likely that that there is no one key influence; rather, there is a diverse range of possible

influences on women’s participation, which include norms, culture, personality and self-

efficacy (as summarised in Appendix 4). There is a need for a research approach that can

recognise both environment and person-based influences pertinent to the DCI context,

and in a manner that helps us understand their complex relationships whislt being

sensitive to gender.

2.5 INITIAL METHODOLOGICAL CONSIDERATIONS

The literature review has raised a number of methodological considerations

regarding how influences can be identified and how they can be understood. This section

introduces six methodological considerations (as listed below). Chapters 3 and 4 present

an in-depth discussion of how these considerations are addressed.

1. Empirical data 2. Non-essentialist approach to gender 3. Multi-level and relational analysis of influences 4. Pluralistic paradigmatic perspective 5. Theoretical understanding of the phenomenon 6. Applied findings

1) EMPIRICAL INSIGHTS

Empirical insights of the phenomenon are required, where these empirical

insights stem from the women themselves. As recognised in the ICT-related literature,

there is a need for research that explores women’s experiences (Adam, Howcroft, &

Richardson, 2002). The broader literature recognises a similar need. An international

Women in Technology 2007 report published by WITI (Women in Technology

International) and reported in an online magazine Eweek notes that “the actual experience

of women working in technology is rarely addressed” (Appendix 17, Item 47).20 A focus

on participants’ perspectives is not because this approach is privileged above others;

rather, it is because insights from aware individuals can offer a certain depth to findings.

Accordingly, a suitable research approach may be one that employs techniques to capture

rich qualitative empirical data from women working in the Australian DCI.

2) AVOIDING ESSENTIALISM

The review of literature suggests that when investigating women’s participation,

there is a need to avoid essentialist assumptions. Concerns have been raised regarding

research that is essentialist towards women (Pini, 2001). Essentialism is the error of

assuming that all instances or cases of a phenomenon or entity are the same (Sayer,

2000). Hence, such essentialism may view gender as an immutable or fixed category,

where all women are seen as having the same characteristics. This is despite the fact that

gender categories are not stable (Connell & Dowsett, 1993), and that the term ‘gender’

has confusing, almost limitless meanings (Risman, 2004). Trauth (2006) noted that

studies underpinned by a positivist paradigm can have an essentialist aspect, particularly

20 http://www.eweek.com/c/a/IT-Management/Report-Techs-Gender-Gap-Widened-by-Uninviting-Workplace/

those studies that investigate women’s traits. In contrast, socialisation theories (often

associated with a constructivist paradigm) recognise that causes are found in the wider

society (Trauth, 2006). However, even socialisation theories can be essentialist if they

focus on differences between men and women (Lewis & Simpson, 2007). Trauth’s

Individual Differences Theory of Gender and IT (IDT) (Trauth et al., 2004) responded to

such essentialism by focusing on the “differences within rather than between the

genders” (Trauth, et al., 2008b, p.26). Trauth et al. (2004) emphasised that women

should not be treated as a homogenous group; rather, individual differences in

experiences must be recognised. The message for this research is that its approach must

avoid essentialism towards gender, or women.

3) MULTI -LEVEL ANALYSIS

As there are a range of potential influences, manifesting across different

analytically distinguishable levels, there is a need for multi-level and relational analysis

of data. Different levels include the individual, group, macro and micro (Ahuja, 2002).

These levels can also be articulated as epidemiological, behavioural and environmental

levels (See Teague’s 1997 study of women working in the Australian ICT industry).

Macro levels can include organisations and structures, and micro levels can

include individual behaviours and perceptions (Klein, Cannella, & Tosi, 1999). However,

there are a number of definitions of ‘macro’ and ‘micro’ (Ritzer & Smart, 2001; Layder,

2006, p.262); ‘micro’ can also involve the individual’s immediate proximal environment

(for example, organisational context), and ‘macro’ the more distal aspects such as global

conditions and wider social structures. Methods used to investigate a phenomenon must

be capable of grasping the relationship between the macro and micro elements of social

life (Layder, 1993).

Cross-level inference (Gerring, 2007)15F

21 may be useful in this regard as this

analytical approach can infer influences manifesting at one level as a result of

investigation at another level. For example, individual level analysis may provide insight

into group and organisation level influences. Although individual women’s accounts of

participation and influences are central, group level inference aims to synthesise

individuals’ insights. Seeking patterns at the group level must, however, avoid an

essentialist stance towards women. As Connell (1987) suggested, a suitable approach

when researching gender is one that conducts research “without collapsing towards 21 Although cross-level inference mostly refers to the inference of variables in quantitative research, Gerring (2007) draws attention to its use in case study.

volunteerism and plurism” on one side, or “categorical and biological determinism on the

other” (p. 61). There is a need to recognise influences at both the individual level and the

‘women’ group level in a manner that is non-essentialist.

Furthermore, a multi-level approach can help reveal those influences less readily

identifiable by observation or analysis of participants’ accounts. In the ICT-related

literature, these less readily identifiable influences have been described as latent and

covert (Roberston, Newell, Swan, Mathiassen, & Bjerknes, 2001); subtle (Webb &

Young, 2005); and implicit in contrast to explicit–where implicit aspects include unstated

assumptions, values, and norms (Weisinger & Trauth, 2003, p.27). These unseen

influences may contribute to what is metaphorically conceptualised as the “glass ceiling”

(Hutlin, 2003; Meyerson, & Fletcher, 2000, p. 127). Recognising less visible influences

is pertinent because influences such as gender discrimination may occur in a manner that

“people don't even notice them, let alone question them” (Melymuka, 2000). However,

the ‘invisibility’ of such influences presents a challenge to those studying them. Perhaps,

this explains Cludt’s (1999) assertion that these influences remain largely unexplored.

Suitable methodological approaches to finding less visible influences include

what Trauth and Howcroft (2006) described as “empirical sensitivity” (p. 282). This

entails the researcher exploring deeper aspects within the data to recognise these unseen

causes; for example, shifting from examining overt discrimination to subtle forms that

produce gender inequity. Another approach is offered by Roos and Gatta (2009), who

suggest that these subtle forms may be understood as the causal mechanisms that

underpin discrimination. Accordingly, methodological approaches that foster multi-level

and relational analysis are required. An exploratory approach must initially keep sight of

the different levels.

4) PLURALISTIC PARADIGMATIC INSIGHTS

There is a need for pluralistic paradigmatic insights. Previous related studies in

the ICT field have utilised positivist, interpretive (associated with social

constructivism16F

22) and critical paradigms. There has been no agreement on which

paradigmatic position offers a foundation for researchers investigating women’s

participation, but rather, there is a debate akin to the ‘paradigm wars’17F

23; this, in turn,

draws attention to the limitations of certain paradigms. For example, studies arising from

22 ‘Social constructionism’ refers to the development of phenomena relative to social contexts, while ‘social constructivism’ refers to an individual's making meaning of knowledge relative to social context. 23 See Clegg and Hardy, 1996 (pp. 5-8) for a succinct review of the ‘paradigm wars’.

a positivists paradigmatic approach may be “essentialist and determinist” (Howcroft &

Trauth, 2008, p. 188); thus, previous research has favoured interpretivist approaches. A

critical paradigm is seen as offering an alternative to interpretive and positivist

approaches (Trauth & Howcroft, 2006; Howcroft & Trauth, 2008).18F

24 Debates indicate

that we can understand women’s participation from different paradigmatic perspectives.

Drawing on the broader sociological literature, Layder (1998) suggested that a

suitable paradigmatic approach may be one that provides a balance between perspectives.

Indeed, the current research climate is conducive to research with a pluralistic approach

to paradigmatic stance (Mir & Mir, 2002); consequently, a choice of method is not

defined by paradigmatic position (Tashakkori & Teddlie, 2003). Kincheloe (2001) went

so far as to claim that the social, cultural, epistemological, and paradigmatic upheavals

and alterations of the past few decades leave rigorous researchers with no choice but to

embrace such an approach. Perhaps, as Smith (2006) noted, it has become evident that a

standard account of the paradigms “suffers from persistent theory-practice

inconsistencies” (p. 2). Accordingly, Section 3.3 proposes Critical Realism as a suitable

paradigmatic foundation.

5) THEORETICAL UNDERSTANDING

There is a need for a theoretical understanding of the phenomenon of women’s

participation in the DCI. As noted earlier, empirical insights are pertinent in light of

there being very little previous research in the DCI context. However, theory can help

explain empirical insights. Trauth et al. (2004) suggested that one of the “research

challenges in studying the under representation of women in the IT field is that of

developing appropriate theory to provide a basis for understanding and explanation about

this gender balance” (p.114). Any “focus on the collection and analysis of empirical data

without the development of an appropriate theory to explain and understand the data”

(Trauth et al., 2004, p.114) may result in the phenomenon being under-theorised (Trauth,

2002). There is certainly the need for a theoretical understanding of women’s

participation in the DCI.

Theoretical understanding may emerge in three ways. First, theoretical

understanding can be a result of empirical data analysis using an inductive logic to

develop conceptual and theoretical concepts. Second, a guiding analytical framework

from previous research can provide a starting point for inductive analysis. As Rowlands

24 A critical paradigm should not be confused here with the critical realist philosophy.

(2005) proposed, a model emerging from a literature review can inform a “loose

conceptual model", and be used as an “initial coding scheme for the qualitative analysis

of data” (p.8). Third, a theoretical understanding can emerge from drawing on existing

models or theory to explain empirical data. This approach has featured in previous

research in the ICT domain. For example, the IDT emerges from empirical studies and

draws on a wide range of other theories including: feminist standpoint theory (FST) and

poststructuralist feminism (Trauth et al., 2009). Similarly, Oswald’s (2008) empirical

study, focusing on stereotypes women encounter due to their gender, proposed the value

of models other than those found in the ICT and gender-related literature; thus, they

suggested that Lent, Brown and Hackett’s (1994) Social Cognitive Career Theory is

relevant. Adam, Howcroft and Richardson (2004) identify that few of the studies in the

ICT domain use theory to explain findings. However, there are several theories and

models that may provide a suitable theoretical perspective, as summarised in Table 7.

Accordingly, Section 3.2 explains how both empirical data and existing theory can be

used to develop a theoretical understanding of women’s participation in the DCI.

Table 7 Theoretical approaches to explaining influences on women’s participation in ICT Name Author/

Year Country Key points Cohort Conceptual/

Empirical/

Individual Differences Theory (IDT)

Trauth, Quesenberry & Morgan (2004)

USA Typology of factors shaping personal characteristics and environmental factors; key concept of ‘Individual Differences’

Women working in ICT

Empirical/ conceptual

Adya & Kaiser (2005)

USA Structural, social and individual differences; modified Ahuja’s model with addition of Trauth’s concept of individual differences

School- aged girls

No empirical application; conceptual (literature review)

Stage-model of barriers

Ahuja (2002)

USA Social, structural influence; focuses on career choice and continuation

Empirical model; no further empirical application

PRECEDE model

Teague (1997)

AU Personal factors and causes within a framework of ‘diagnosis’; for example, social factor

Conceptual (health domain theory); no empirical application

Webb & Young (2005)

AU Descriptive/conceptual categories, including: - supermum - serendipity - culture

Empirical model; no further empirical application

Conceptual framework for studying gender and IS

Wilson (2004) UK Combines insights derived from: (1) gender and computing, concerning the differences and inequalities of development and use of IS; (2) gender and society concerning the existence of gendered spheres; (3) studies of gender and organisations concerning the social division of labour; and (4) gender and technology, concerning the masculinity of technology culture.

Women working in IS

Conceptual model, no further empirical application

6) APPLIED FINDINGS

Academic, industry and government literature suggests that a range of strategies

have been employed to foster both girls’ and women’s participation in the ICT industry

and, more recently, in the DCI. Table 8 identifies several exemplars of these strategies,

and further examples are provided in Appendix 3. The majority of these strategies target

school-aged females, and are based on the notion that encouraging young women may

translate to a choice to pursue technology-related careers.

There are fewer strategies to foster women’s participation once they are in the

workforce. In Australia, a small number of ICT organizations, including IBM and DELL,

propose best practice–or at least better practice–initiatives.25 These strategies or

interventions are not usually rigorously implemented, and there are few evaluations of

the efficacy of the intervention (Craig et al., 2011). Strengthening these initiatives may

involve the application of theoretical knowledge. Trauth (2011) proposed that there is a

“tremendous need for theoretically-informed interventions in order to properly address

real-world problems” (p.7). There is a need for “a theory whose constructs can help to

explain the gender imbalance in such a way that appropriate interventions based upon

them can be developed” (Trauth et al., 2009, p. 493). In other words, a research

approach based on empirical data and theoretical insights could offer a way to understand

appropriate initiatives.

Table 8 Examples of strategies/initiatives to foster women’s participation (a full table with further information is included in Appendix 3) Strategy /Initiative Industry1 Area 2 Country WINTIT ICT Education (Graduate) UK, AU Digital Divas ICT Education (K-12) AU

WIT (WIT, 2006) ICT/DCI Industry UK, AU

IGDA mentorship scheme DCI (Games) Industry USA 3

Technology Takes you Anywhere ICT/DCI Education (K-12) AU

Diploma of Multimedia (Women) 4 DCI Education(Vocational) AU

Girl Geek Coffee Club ICT/DCI Education(Graduate) AU Geek Girl Dinners ICT/DCI Education, Industry UK 3 Go GURL ICT Education (K-12) AU Go Girls, Go for IT ICT Education(Graduate) AU Women in Games DCI (Games) Industry USA

1 Industries often overlap, 2 does not closely examine sources of funding 3 international strategy 4 last offered as a gender specific program in 2007

25 Dell Australia proposed a pilot ‘Women in IT mentoring’ program involving five ICT firms and aimed at improving the career progression opportunities for women in the sector. http://www.minister.dcita.gov.au/media/speeches/participation_summit_2005_opening_address

2.6 CHAPTER SUMMARY

The literature review, so far, has introduced the phenomenon under study, that of

women’s participation. The review has also introduced the context of the phenomenon–

the Australian Digital Content Industry (DCI) (Section 2.2).

Section 2.3 reveals that even though the DCI is a growing industry, women

remain under-represented in the industry, both internationally and in Australia. The

historical and ongoing under-representation leads to the research problem: Why do

women participate in the Australian DCI?

A number of implications for this study emerge from the literature review. The

number of women working in multimedia related roles and those in games related roles

appears to vary, with there being more women in the multimedia sector. This may be a

result of the more recent emergence of the games sector. The implication for this study is

that the experiences of both women working in multimedia and games must be

investigated.

There also appears to be a segregation of the types of roles, with there being

fewer women in technical production roles. Therefore this study aims to investigate the

experiences of women employed as interactive content creators (as per the ABS

definition), who work in a technical role. Accordingly, this investigation focuses on

women working as interactive content creators in technical roles within multimedia and

games development organisations.

Section 2.4 identified the paucity of academic research regarding women’s

participation in the DCI, thus further establishing the need for this investigation. Both the

ICT and DCI literature identify a diverse range of possible influences on women’s

participation. The first research question thus emerges and asks: What are the influences

on women’s participation in the DCI?

The review identifies that there are a wide range of possible influences on

women’s participation and these include environmental characteristics (such as the

stereotypes women face) and person characteristics (such as personality). These

influences may vary across different contexts (for example, multimedia or games

organisations) and occupational roles (for example, design or programming).

Consequently, there is a need for a research approach that can recognise both

environment and person-based influences pertinent to the DCI context, and in a manner

that helps us understand their complex relationships whislt being sensitive to gender.

Section 2.5 raised a number of initial methodological considerations regarding

the identification and understanding of influences. Accordingly, the second research

question asks: How can the influences on women’s participation be understood? Previous

research, or the gaps in this research, indicate that there is a need for empirical data as

there is little in the way of women’s experiences. In addition, there is a need for

theoretical and explanatory insights to understand this data. The investigation must aim

for a non-essentialist approach to understanding women’s experiences. A pluralistic

paradigmatic approach may be of benefit in light of there being complex, multi-level

interactions between possible influences.

As the following chapter (Chapter 3) proposes, these methodological

considerations can be addressed, in part, by adopting three frameworks with which to

analyse the empirical data. It is proposed that these three frameworks offer a reliable and

holistic way to investigate the phenomenon under study.

Chapter 3. Ways to understand influences ‘Women’s Participation in the Australian Digital Content Industry’ Author: Anitza Geneve Year: 2013 Page 52 of 401

Chapter 3: Ways to understand influences

As Chapter 2 established, although there is little research regarding the influences

on women’s participation in the DCI context; previous academic literature indicates that

there are a diverse range of influences on women’s participation in the ICT industry.

There is a need to identify the influences women experience within the DCI context.

However, is not enough to identify a disparate range of influences; rather, it is vital that

influences be understood in a unified manner. Accordingly, this chapter proposes that

three frameworks (as introduced in Chapter 1, Table 2) can assist in the identification and

understanding of influences.

Section 3.1 considers the value to the research process of Framework 1, an initial

guiding analytical framework comprised of three categories: 1) Environment, 2) Person,

and 3) Interaction. Framework 1 acknowledges that there is likely to be an interaction

between both the environmental and person characteristics , a key point emerging from

the literature review.

Section 3.2 considers the value of Framework 2, a theoretical framework

comprised of several Human Agency theories. Framework 2, the Human Agency Multi-

Theory Scaffold (HAMTS), acknowledges that although there are likely to be a range of

environmental characteristics, that the individual plays an important role in their

participation. Framework 2, is a specific example of a Multi-Theory Scaffold (MTS). The

MTS offers a research approach for using four different types of theory to scaffold or

guide the data analysis and theorisation process. Both the MTS and HAMTS models

emerge as an early research contribution.

Section 3.3 considers the role of Framework 3, a philosophical or ontological

framework based on Critical Realism. Framework 3 offers the philosophical

underpinning of the research and ultimately a way to explain the phenomenon under

study.

It is proposed that individually these three frameworks assist in both identifying

and understanding influences; this, in turn, helps to address the research problem: Why

do women participate in the Australian DCI? As this chapter notes, these are not three

disparate frameworks, rather in contrast, there is a strong theoretical synergy between the

three frameworks.

3.1 FRAMEWORK 1: ANALYTICAL MODEL (E, P, I)

The review of previous research in Chapter 2 raised several methodological

considerations, including the need for a research approach that values multi-level

analysis of empirical data to identify influences that manifest at different levels; for

example, at the person and environment levels. In response, this section proposes an

analytical framework for the analysis of empirical data that is comprised of three

categories: 1) Environment, 2) Person, and 3) the Interaction of the Environment and

Person. The following is an explication of how related literature supports the

conceptualisation of Framework 1.

A gap exists in the research for a model that can help explain women’s

participation. Ahuja (2002) states that: “No definitive work has presented a model that

explains the role of women in the field of IT in a comprehensive manner” (p. 21). One

important theoretical perspective within the ICT-related research is Trauth’s Individual

Differences Theory of Gender and IT (IDT) (Trauth et al, 2004;Trauth et al., 2005;

Trauth, 2002, 2006; Trauth et al., 2009). IDT draws attention to the “causes of gender

under-representation” that women working in the information technology profession may

experience (Trauth et al., 2004, p. 114). IDT may provide a suitable starting point for any

study interested in women’s participation in the ICT industry.

IDT presents a conceptual framework that recognises three main categories of

influences: 1) personal, 2) shaping and influencing factors, and 3) the environmental

context. As Table 9 indicates, for Trauth et al. (2004), environmental influences include

cultural attitudes towards women; personal influences include identity characteristics;

and shaping and influencing factors include educational and life experiences of the

individual. IDT recognises that these influences are constructed at the “individual level of

analysis” and, thus, the differences in experiences of participation are in part due to the

individual differences among women (Trauth et al., 2004, p.114).

Table 9 Conceptual framework offered by IDT (Trauth et al., 2004; Trauth, et al., 2005) Category of factor Construct Example Personal Data Descriptive data about the participant

Demographic Data

Age, Ethnicity, Gender, Nationality, Race, Religion, Sexual Orientation

Lifestyle Data Children, Family Background, Family Work Background, Spouse/Partner

Workplace Data Career Characteristics, Industry Type, Job Title, Technical Level, Type of IT Work

Shaping and influencing factors; personal characteristics and influences experienced by that individual, such as: early experiences with computing, role models, mentors and significant others, significant life experiences

Personal Characteristics

Education, Interests and Abilities, Personality Traits, IT Identity, Gender Identity

Personal Influences

Exposure to Computing, Educational Experiences, Life Experiences, Role Models and Mentors

Environmental Context; provides context within which the person’s responses are situated

Cultural Attitudes and Values

Attitude Toward Women, Women Working, Women Working in IT; Academic Attitudes toward Women (In General, in IT); Workplace Attitudes Toward Women (In General, in IT)

Geographic Data Location, Population, History Economic Data Employment Overall, Information Economy

Employment

IDT suggests that person factors and environment factors are an influence on

women’s participation. IDT provides a third category that appears to overlap the

environmental and person factors; ‘shaping and influencing factors’ are a “combination

of personal characteristics possessed by the individual and influences experienced by the

individual” (Trauth et al., 2004a, p.118). This third category in IDT importantly draws

attention to the interaction of influences. Section 3.1.3 further discusses the need to focus

on the interaction of influences.

Although, the IDT appeared to offer a useful framework with which to

understand the influences on women’s participation in the DCI, an original guiding

analytical model is proposed for three key reasons. First, the categorisation of several

sub-categories of the IDT model is questioned. For example, Trauth et al., (2004)

position the ‘workplace data’ construct under the personal category. However, it may be

considered that workplace characteristics may also be relevant to environmental context.

Thus, the conceptualisation of the IDT categories would have required further

clarification prior to application in this research. Second, IDT does not explicitly focus

on interaction as an analytical category, or specify the manner in which this interaction

may be conceptualised or explained; this is despite the fact that Trauth et al., (2005)

clearly suggest there is a “benefit” in “considering the interaction between constructs” (p.

31) such as gender and geographical location. As Section 3.1.3 emphasises, an

investigation of the interaction of influences is required. Third, as the phenomenon of

women’s participation in the Australian DCI is under-researched, an exploratory research

approach is beneficial. Framework 1 is thus comprised of three categories: 1)

Environment, 2) Person, and 3) Interaction between these two. As illustrated in Figure 4,

the framework is comprised of three broad categories; this is in contrast to the detailed

categories and sub-categories offered by IDT. There is, however, a resonance between

the categories in Framework 1 and IDT, even if the framework does not explicitly draw

on IDT: both consider the environment and person influences. Framework 1 is used to

analyse empirical data.

Figure 4: Framework 1: Analytical Framework comprised of three categories (E, P and I)

Further research supports the relevance of the three categories in Framework 1.

As summarised in Table 59 in Appendix 4, the existing literature provides multiple

examples of influences in both the Environment (E) and Person (P) categories. These

influences include:

1. Environment influences, which include structural and social characteristics [Structural characteristics include those of the industry (such as long hours), and artefacts (such as resources, computers, and digital products); social characteristics include cultural, historical, media, and family.]

2. Person influences [These include identity, behaviours, personality traits (such as confidence), skills, and perceptions.] With regards to the third category (I), although previous research emphasises that

an investigation of women’s participation must consider the interaction between

influences, few studies have explored women’s participation by explicitly focusing on

the interaction of influences. Influences identified in previous research are now

considered with reference to the three categories in Framework 1: Environment (E),

Person (P) and Interaction (I) between E and P.

Analysis

Empirical data

1. Environment 2. Person

3. Interaction

3.1.1 ENVIRONMENT (E) INFLUENCES

The first category in Framework 1 is the Environment (E). Previous research

identifies a wide range of environmental characteristics, including structural and social

characteristics, which may influence women’s participation in the DCI; these structural

and social characteristics include workplace practices, culture, social practices, and

resources within the environment.

Women’s under-representation has been attributed to specific structural

characteristics, such as work patterns. For example, the culture of working long hours

has been recognised in both the DCI (Gill, 2002) and ICT industry (Ahuja, 2002;

Griffiths, Moore, & Richardson, 2007). In DCI organisations, these long hours often

manifest near project completion, known as “crunch-time” (Consalvo, 2008; Jenkins &

Cassell, 2008, p. 11). In the USA, the games industry has been criticised for work

practices that require long hours, and even unpaid hours, as evident in the 2006 EA

‘spouse’ case in the United States26.21F Such work patterns may explain why women

working in Australian ICT value a flexible work environment (Queensland Government,

2006a)22F27. However, Roan and Whitehouse’s (2007) found that in the Australian DCI

context that work hours were regular and reasonably flexible. Therefore, even though a

characteristic of the environment, such as long hours, is widely cited as an influence, a

suitable research approach must remain open as influences may vary among different

environmental contexts.

Although many of these environmental characteristics may influence both female

and male DCI workers, certain characteristics may be particularly relevant to women’s

participation. In the ICT literature, Ahuja (2002) identifies a range of structural barriers

pertinent to women, including occupational culture, lack of role models and mentors23F

demographic composition and institutional structures. Prescott and Bogg (2010) noted

the lack of females, especially females with families, in the gaming industry in the UK

games’ industry context. Access to childcare has been identified as influencing women’s

participation in ICT (Queensland Government, 2006b). However, childcare access may

be an influence on the participation of either males or females. To suggest that childcare

access is an influence only for women is to take an essentialist approach to gender.

Essentialism is an assertion that there are fixed, unified and opposed female and male

natures (Trauth et al., 2004, citing Wajcman, 1991). Accordingly, a suitable research 26 Stemming from a spouse’s posting to a blog, software engineers won a $14.9 million settlement from Electronic Arts in 2006 for unpaid overtime (See Item 37, Appendix 17: Secondary sources of data). 27 partICipaTion National ICT Skills Summit 28 Agosto, Gasson and Atwood (2008) suggest that a mentor differs from a role model as the former offers personal contact.

approach must initially consider a wide range of environmental influences, not only those

that appear (on the surface) to be most relevant to women.

Environmental characteristics that may influence women’s participation also

include social characteristics of the environment. Trauth’s framework (See Table 9), for

example, identifies “socio-cultural influences” (Trauth, 2002, p. 98) such as attitudes to

women. Other researchers identify social expectations, the work–family conflict,

informal networks, and gender stereotypes that surround women (see Ahuja, 2002).

Stereotypes include those that suggest women do not have an affinity with technology

(Wajcman, 1991). These stereotypes can lead to women feeling as though they are not

the ‘norm’ and that they are “the odd girl out” (Trauth, 2002, p.98). However, recent

research suggests that there may be a ‘possible shift’ occurring in younger people’s

perceptions of gender stereotyping of ICT-related skills (Trauth, Joshi, Kvasny, Chong,

Kulturel, & Mahar, 2010). There are also occupational stereotypes associated with

workers, where DCI occupations are seen as ‘cool’ (Gill, 2002), and ICT roles as being

‘geeky’ (Beekhuyzen & Clayton, 2004). However, there is a difference between the self-

perception of IT workers and their public perception (Courtney, Lankshear, Anderson, &

Timms, 2009). These social influences manifest within the organisational and wider

social context, including the mass media (Barker & Aspray, 2006). Social characteristics

at the organisational level can include organisational culture, which has been perceived

as ‘chilly’ (Roldan, Soe, & Yakura, 2004), ‘competitive’ (Webb & Young, 2005) and

‘hostile’ (Griffiths & Moore, 2010). Social characteristics of the environment are worth

exploring because participation is a social phenomenon, often involving other people.

Although the review of existing literature draws heavily on the ICT literature,

there are some unique aspects of the DCI environment that may influence women’s

participation, such as the nature of the digital products produced. Games are becoming

ever more increasingly popular amongst women in Australia (Brand, Borchard, &

Holmes, 2009). However, women’s under-representation has been attributed in part to

the lack of appeal or suitability of the digital products the DCI produces, where computer

games sexualise female characters (Beasley and Standley, 2002), and target content at

males (Pham, 2008). 24F

29 These game products can “serve as an entry point to the culture

of computing and information technology” (Brunner, Bennett, & Honey, 1998, p. 41).

Fullerton, Fron, Pearce, and Morie (2008) propose “that women would aspire to be game

29 Games such as Grand Theft Auto have received considerable media coverage for their violent and negative portrayal of women.

designers more frequently if there were more games available that they enjoyed” (p.

165). Thus, resources or artefacts in the environment, such as computer games, may

influence women’s participation and warrant investigation.

3.1.2 PERSON(P) INFLUENCES

The second category in Framework 1 is the Person (P). Person characteristics

may influence women’s participation. In the ICT context, Trauth’s IDT category

‘personal data’ draws attention to a “combination of personal characteristics” of the

individual (Trauth et al., 2004, p. 118), which include demographic characteristics such

as age and gender (as indicated in Table 9). In addition, the ‘shaping and influencing

factors’ that Trauth et al. (2004) propose include personality traits, interests and abilities,

and gender and occupational identity. Other research in the ICT domain suggests

personal influences may include behaviour (Teague, 1997), and cognitive aspects such as

self-efficacy (Adya & Kaiser, 2005; Betz & Hackett, 1997). The role of self-efficacy for

women working in the ICT industry continues to draw attention from researchers

(Kvasny, Joshi, & Trauth, 2011; Joshi, Kvasny, McPherson, Trauth, Kulturel-Konak, &

Mahar, 2010). Thus, existing research suggests that there are a range of person

characteristics that may influence women’s participation.

As noted earlier, there is a need for a non-essentialist approach to gender when

researching women’s participation. Trauth et al. (2004) highlight the importance of the

“individual differences” among women (pp.114-115).25F

30 Recognising individual

differences acknowledges both the different circumstances women face (such as the

environmental contexts) and women’s individual capabilities and responses. “[Women]

do not all experience the same influences, nor do they all respond in the same way”

(Trauth, 2002, pp. 114-115). Trauth et al. (2009) suggest that it is these individual

characteristics or individual differences that can explain why previous studies present

“contradictions” (p. 482) regarding influences. In other words, the influence of

environmental characteristics may be dependent on the individual’s response. However,

the broader literature notes that characteristics such as individual consciousness are often

overlooked when exploring inequity in Australia (Greig, Lewins, & White, 2003, p. 3).

Accordingly, the person category must recognise the role the individual woman plays in

her participation.

30 Trauth et al. (2004b) note that IDT draws on empirical studies and sources that discuss “individual differences”; these include Jennings (1941) who focused on relations between individuals, and McCauley and Thangavelu (1991) who focused on stereotypes.

3.1.3 INTERACTION (I) BETWEEN THE ENVIRONMENT (E) AND PERSON(P)The third category in Framework 1 is Interaction (I); that is, the interaction of the

environment and the person. Research in related domains suggests the importance of

investigating the interactions among influences. Sociological literature emphasises the

need to understand the socially bounded interactions between the environment (for

example, other people, social practices, norms and beliefs) and the individual (Layder,

1998). In the domain of career theory, Savickas (2005) states that it is not enough to

identify the person (P) and environment (E) influences, or P-E; there is also a need to

focus on the “dash” (p.45). In the ICT context, Ahuja (2002) proposes that “social and

structural factors as well as their interactions will result in turnover of women in IT”

(p.20), and that in order to develop a “rich understanding of IT careers, it is crucial

[emphasis added] that interactions among these factors be considered” (p. 22). In

Australian-based research, von Hellens and Nielsen (2001) propose that there is a need to

focus on the “relationships” between the “complex mixture of factors” (p. 52). It is the

“understanding of the relationships and interaction of all these factors that holds the key

to addressing the declining interest in ICT education and work” (von Hellens, Clayton,

Beekhuyzen, & Nielsen, 2009, p. 214).

A handful of empirical studies regarding women’s participation in the ICT

industry recognise the relevance of interactions between influences. For example, Adya

and Kaiser (2005) recognise the interaction between the individual and the ethnic, social

and structural factors that influence career choice. However, few studies investigating

women’s participation explicitly consider ‘interaction’ as an analytical object. Trauth’s

IDT (Trauth et al., 2004, 2005) comes closest, but does not offer an explicit focus on

interaction; rather, the category ‘shaping and influencing factors’ implies interaction

between certain individual and environmental influences. Trauth et al. (2005) suggest

that IDT draws attention to interactions among technological, individual and societal

forces (p.1). These forces include factors such as individual (for example, parenthood)

and organisational (for example, policies and norms), and gender and geographical

location. An example of such interaction may be evident in an influence that Trauth et al.

(2004) identify as ‘exposure to computing’ (categorised within the ‘shaping and

influencing factors’ category), where access to computers (environmental influence) may

foster a person’s aptitude for technology (person influence). The implication for this

investigation is the need to identify the Environment (E) and Person (P) influences that

may contribute to women’s participation and, importantly, to also focus on understanding

the Interaction(I) of these influences.

There are several strengths in focusing on the interaction between the

environment and person. First, a focus on the interaction among influences may address

criticisms that previous studies in the ICT domain face: that they offer a “wide range of

fragmented and difficult to holistically understand findings” (Quesenberry, 2006, p.343).

Ahuja (2002) argues that it is “not sufficient to examine these factors in isolation from

one another, as the IS literature on gender has done so far” (p. 22). Focusing on

interaction provides a way to link the seemingly disparate influences across several

levels such as structural, social, and individual. A second benefit is that an analytical

focus on the interaction recognises that both the environment and the person contribute to

an individual’s participation. Neither the environment nor the person are held entirely

accountable; both play a role. For example, an influence such as gender stereotypes may

manifest in the environment as cultural norms and stereotypes and, in the individual, as

second order expectations (Webster and Whitmeyer, 1999). However, the environment

may provide role models that help challenge such stereotypes or the individuals

themselves may choose to challenge them and transform them in so doing. Thus, a focus

on interaction may help tell a more holistic story of women’s participation.

Focusing on interaction also raises the concept of causality. Indeed, in applying

IDT, Trauth, Quesenberry, and Yeo (2008) use terminology that implies effect; for

example, ‘to shape’. This term implies that when aiming for a “deeper examination” of

factors, there is a need to recognise “underlying causes of different responses to them”

(p. 27). In gender-related studies, a similar call has been made to examine causal effects

by investigating mechanisms; that is, any “causal explanation must address how a

relationship came about” (Reskin, 2003, p. 16). An approach that considers interaction

can illuminate causal effects and can thus add to an overall explanation of influences and

the phenomenon of women’s participation.

A consideration of causality may also illuminate temporal aspects of the

phenomenon. Previous research identifies a temporal aspect to influences. In the ICT

domain, Ahuja (2002) identifies that certain groups of factors are, for example, more

influential at different career stages. Gurer and Camp (2002) metaphorically describe the

gradual decline in women’s participation as they progress along a school, education and

career pathway as the “shrinking pipeline”. The causal effect of influences may vary

over time. Accordingly, a focus on the temporal aspect of interactions, or of the

temporal context of interactions, can help illuminate influences.

3.1.4 SUMMARY

The review of previous research indicates that the influences on women’s

participation may be categorised into the two broad categories of Environment and

Person. However, a consideration of the interaction between the environment and person

is crucial. Accordingly, the research approach has identified an analytical framework

comprised of three categories of influences: 1) Environment (E), 2) Person (P), and 3)

the Interaction between those two (I). This first framework, Framework 1, serves as an

initial guiding analytical framework for the analysis of empirical data, where the aim is

to identify and understanding the influences on women’s participation.

Framework 1 can help address several of the methodological considerations

raised in Section 2.3, as it encourages multi-level analysis of both macro and micro

influences (such as structural and individual characteristics). Section 2.3 also raised the

need to develop a theoretical understanding of the phenomenon of women’s

participation. Although this section has introduced possible theoretical approaches from

the DCI or ICT literature–for example, Trauth’s IDT, certain limitations have been noted

regarding the suitability of these approaches. Previous research has not had an explicit

focus on the interaction between influences. Accordingly, the following section looks to

the broader literature for a suitable theoretical perspective that focuses on the interaction

between the environment and person, and that can provide an approach through which to

understand the phenomenon of women’s participation.

3.2 FRAMEWORK 2: THEORETICAL (HUMAN AGENCY THEORY)

Section 2.3 identified the need for a theoretical understanding of the phenomenon

of women’s participation in the DCI. In response, this section presents three arguments:

first, that existing theory is useful in the research process; second, that multiple theories

are warranted for the research problem at hand; and third, that Human Agency theories

offer a suitable theoretical perspective. Section 3.2.1 considers the methodological

concerns of using existing theory in exploratory research, and introduces the approach of

using theory as a “scaffold” (Walsham, 1995b; Layder, 1998). Section 3.2.2 provides an

original contribution by proposing a Multi-Theory Scaffold (MTS)–a model guiding the

choice of more than one theory in a scaffold. Section 3.2.3 proposes that Human Agency

theory provides a suitable theoretical perspective from which to understand women’s

participation. Building on the argument in Section 3.2.1–that more than one theory is

helpful to the research process–it is proposed that there are four agency theories that can

be used to understand the influences on women’s participation. Thus, Section 3.2.4

presents the four Human Agency-related theories that comprise Framework 2. This

framework is referred to as the Human Agency Multi-Theory Scaffold (HAMTS) from this

point on.

3.2.1 THE ROLE OF THEORY

Existing theory can be an important tool in an inductive research process

(Walsham, 1995b; Layder, 1998). However, in practice, the use of existing theory is

contentious and presents the researcher challenges. Indeed, Sayer (1992) states that “any

serious consideration of method in social science quickly runs into basic issues such as

the relation between theory and empirical observation and how we conceptualise

phenomena” (p. 45). Thus, a critical question arises for this investigation: How can

existing theory be utilised in a largely inductive research approach? Inductive referring to

what Morse (cited in Tashakkori & Teddlie, 2003) calls the “thrust of discovery” (p.193,

Accordingly, this section considers the role of theory. There are several strengths

in using existing theory as a way of understanding the influences on women’s

participation. This argument can be built on a consideration of Dobson’s (1999)

suggestion that there are four distinctive approaches for the use of theory in an

interpretive, in-depth case study.

1. No theory: grounded theory, for example 2. Single Theory: where a researcher has an expert understanding of the theory

(Alvesson, 1996) 3. Multiple Theories: where theory is used as a “scaffold”, to be discarded when

no longer needed (Walsham, 1993; 1995b) 4. Context-dependent use of theory: theories of a substantive nature

Dobson’s observation implies that when considering the use of theory, there are

four questions to be considered by the researcher: 1) Will any theory be used at all? 2)

Will one or more theories be used? 3) How can theory be used? and 4) Which theory

should be used? What follows is a response to these four questions.

In response to the first question–Will any theory be used at all?–there are benefits

to be considered. Researchers within the ICT domain suggest that existing theory may be

useful in a variety of ways; for example, it may: focus, guide, frame and stimulate

theoretical sensitivity by providing concepts and relationships to be ‘checked out against

actual data’ (Rowlands, 2003); offer a crafting tool, such as Giddens ST (Shoib,

Nandhakumar & Jones, 2006); provide a sensitizing device (Duberley, Mallon & Cohen,

2006); offer a new voice and device to re-author experiences (Alvesson & Kärreman,

2007); help in sensitisation (Walsham, 1995b), analysis (Jones, 1999; Rose, 1998) and

theorisation (Jones, 1999); offer a theoretical structure (Stake, 1995) and a set of

dispositions or thinking tools (Swartz & Zolberg, 2004). Thus, existing theory can play a

range of roles in research.

In response to the second question–Will one or more theories be used?–an

argument is put forward for multiple theories. This is because there may be a need for

more than one theory due to the complexity of the phenomenon under study. Previous

studies in related fields suggest that there are limitations in drawing on a single theory. In

the career development field, Super (1992) contends that no one theory in itself is

sufficient to adequately address complexity. In the ICT field, Trauth et al. (2004)

emphasise that “one of the research challenges in studying the under representation of

women in the IT field is the lack of sufficient [emphasis added] theory to provide a basis

for understanding and explanation about this gender imbalance” (p. 114). Although

‘sufficient’ may refer to the paucity of theories that are available to help understand the

phenomenon of women’s participation, it may also be suggestive of a limitation in

relying on a single theory to understand a complex phenomenon. Trauth et al. (2008b)

suggest that “given the complex nature of human behavior, perhaps the most effective

social theories are those that have combined several concepts and/or frameworks in order

to understand social phenomena” (p. 11).

Multiple theories can guide a study or offer a theoretical explanation. Rowlands

(2005) recommends that empirical research needs to be “guided by (or at least informed

by) one or more [emphasis added] social theories” (p.87). Walsham (1995b)

recommends using multiple theories when conducting case study, not as rivals but to

complement each other. Kelle (1995) emphasises that an approach that would use

different and even competing theoretical perspectives on the same data is preferred over

an “emergence” of theoretical notions from the data. Thus, there are benefits in drawing

on multiple existing theories.

In response to the third question–How can theory be used?–there is a need to

recognise that the use of theory varies, particularly within qualitative research (Creswell,

2007; Tashakkori & Teddlie, 2003, p.119). Variation surfaces with respect to “when”

theory should be introduced (Creswell, 2003, p.134), and a researcher must consider

when to most appropriately include theoretical perspectives (Andersen & Kragh, 2009).

Some guidance has been provided for interpretive case studies, where it is suggested that

theory may be used: as an initial guide to design and data collection, as part of an

iterative process of data collection and analysis (Walsham, 1995b) and as a final product

of the research (Eisenhardt, 1989).

One approach to using existing theory is employing theory as a “scaffold”

(Walsham, 1995b, p. 76; Layder, 1998, p.150). In using theory as a scaffold, the aim is

not to make the data fit to the theory; rather, the theoretical scaffold may be

progressively removed, if and when the research no longer requires it (Walsham, 1995b).

This use of theory aligns to research that aims to balance empirical insights and prior

theoretical knowledge, perhaps to avoid “theoretical over determination and empirical

under determination” (Giddens & Turner, 1988, p. 25). The use of theory in this manner

is not new; indeed, James noted (in 1907) that theory serves as an instrument and not

“answers to enigmas” (James, 1907, p. 46). More recently, Walsham (1993; 1995b) used

theory as a scaffold to develop questions used during data collection. Examples of other

researchers using existing theory as a scaffold include Orlikowski and Robey (1991) who

drew on Giddens’ Structuration Theory (ST) (Giddens, 1984).

Theory as a scaffold does not require that an entire theory be applied; rather,

select concepts can be used as orienting devices to impose meaningful patterns in a

provisional manner (Layder, 1998, p. 109). Indeed, some theories, such as Giddens’ ST

(1984), may be “too complex, diverse and alien to be adapted wholesale”, and “relevant

concepts must be selected and adapted into theoretical frameworks” (Rose & Scheepers,

2001); this is done by using what Jones (1999) describes as a “pick and mix” process.

Even Giddens (1984) recognises that the use of concepts from theories may be critical

and “selective” (p.43). Thus, employing existing theory as a scaffold entails the select

use of relevant concepts.

There are some limitations to consider when drawing on select concepts from a

theory or theories. Layder (1998) recognises the possible “damage” caused when the

explanatory power of a theory may be impacted by applying single concepts incorrectly,

or by the “fragmentary and partial appropriation” (p. 97) of select concepts. Similarly,

Craib (1992) has argued that by drawing on several theories, one may lose “vital

dimensions of each of them and they cease to do the theoretical work” (p. 63). This is a

criticism Craib has specifically directed towards Giddens.

While the researcher must be aware of these limitations, there are also many

benefits of employing existing theory as a scaffold when developing and elaborating

theory in empirical studies. Layder’s Adaptive Theory (AT) (1993, 1998) articulates these

benefits. AT builds on the principles of Critical Realism to recommend an approach that

entails a blend of inductive and deductive logic. In contrast to traditional inductive

research processes, where reasoning is grounded in empirical observations, Layder

(1998) sees general and substantive theory as additional “raw material” for theorising (p.

163); thus, the researcher iterates between existing “prior” theoretical materials (p. 166)

and emergent empirical data to develop theory, and “theorising is a continuous rather

than discrete aspect of the process” (p. 174). The exact form AT takes can vary, and it

requires a flexible approach to theoretical and empirical resources and methodological

strategies and techniques (Layder, 1998, p. 158). However, a key aspect of AT is that

existing theories do not function as a “ready made” explanation of findings (Layder,

1998, p. 23). Rather, Layder’s (1998) AT recognises that the “ransacking” of general

theories (p. 164) stimulates analytical thinking so that “good extant ideas” are not

“wasted” (p. 114), and that concepts from existing theory offer orientating devices which

serve as a means of “cranking up the process of theorising–either by elaborating on

extant theory or by generating theory in relation to research evidence” (Layder, 1998,

pp.23-24). Thus, employing existing theory as a scaffold entails the use of both empirical

data and existing theory for new theory development.

The first three questions proposed in Section 3.2 have now been addressed to

establish: that theory is useful, that more than one theory is useful, and that theory can be

used as a scaffold. In regards to the fourth question–Which theory should be used?–there

is, of course, the need to use those theories appropriate to the research problem, and not

simply those which are readily on hand, or popular. To answer this fourth question

regarding theory, therefore, there is the need to consider two aspects: the value of

different types of theories (to be discussed in Section 3.2.2); and the identification of

theoretical perspective suitable for the problem at hand (to be discussed in Section 3.2.4).

3.2.2 MULTI -THEORY SCAFFOLD (MTS) This section considers how different ‘types’ of theory can be used as a scaffold to

benefit the research process. Section 3.2.4 will further extend the argument here by

illuminating the value of employing different types of agency theory in a scaffold.

Although there have been several significant papers regarding the value of using

a “scaffold” (Walsham, 1995b, p. 76; Layder, 1998, p.150), there are few guidelines on

the methodological process, particularly if one were to draw on more than one theory. In

contrast, when considering theory 'development’, there is a clear understanding that

different types of theoretical outcomes offer different explanations. Established

frameworks that recognise such levels of theorising include Llewellyn’s (2003) five ways

of theorizing, and Gregor’s (2006) five types of theory as an outcome of a study.

Llewellyn (2003) does not specify different types of theory, but rather, ways of

theorising to include: 1) metaphor, 2) differentiation, 3) conceptualisation, 4) context free

theorising of setting, and 5) context free grand theorising. Llewellyn (2003) does,

however, suggest that there are differences in the types of theories one can develop from

empirical research by making the distinction that grand theory emphasises structural

conditions (p. 677). Gregor’s (2006) five types of theory include: 1) theory for analysing,

2) theory for explaining, 3) theory for predicting, 4) theory for explaining and predicting,

and 5) theory for design and action. Gregor (2006) provides a clear indication that

different types of theories may be developed. However, when drawing on theory as a

scaffold, there is no framework to offer guidance on the choice of types of theories.

Dobson (1999) recognises the limited guidance in regards to the choice of

theories to be used as a scaffold: “A scaffold is a useful metaphor for theory use but it

does not help in defining which theories to use” (Dobson, 1999, p. 264). There is,

however, some agreement that different types of theories are useful, especially for further

theorising (Layder, 1998; Danermark et al., 2002). Layder (1998) specifically notes the

value of general theories, middle range theories and substantive theories. Layder (2006)

differentiates that general theory (such as Giddens ST) is concerned with generic

theoretical problems, whereas substantive theory applies to specific empirical areas (p.

180). Danermark et al. (2002) distinguish four types of theories which may be useful

when drawing on theory from an abductive logic: 1) meta theory, 2) normative theory

(the way things ought to be and focusing on moral, political and ideological issues), 3)

descriptive theories (which describe fundamental properties), and 4) ordering or

conceptual frameworks. Although both Layder (1998) and Danermark et al. (2002) make

distinctions regarding the value that different types of theory bring to research, neither

explicitly states if or how one or more theories may be used in a manner that harnesses

different perspectives. In response, an original contribution is proposed in the form

of a conceptual model: the Multi-Theory Scaffold (MTS).

A MTS is an analytical research tool to help the researcher consider the choice of

different types of theory for inclusion in a scaffold. In agreement with Danermark et al.,

(2002) the MTS recognises that all theories comprise “abstractions” (p. 126). Different

types of theory offer different types of abstraction. For example, general theories differ

from other theories because of their “degree of abstraction and thus their apparent

‘distance’ from empirical matters” (Layder, 2006, p. 95). Danermark et al. (2002)

suggest that middle range theories may “be abstract enough to enable them to be applied

to different social phenomena, and concrete and specific enough to permit testing against

empirical data” (p. 126). In contrast to the highly abstract conceptualisations of meta

theories (Gregor, 2006), operationalised theories offer a focus on specific measurable

variables. Raduescu and Vessey (2009) propose that the use of domain-specific theories

may even influence methodological approaches. Thus, the MTS is comprised of four

types of theory, which may foster deeper insights, as different types of theories offer

different levels of abstraction. As Table 10 presents, theories are categorised according

to the level of abstraction they offer: 1) meta, 2) critical, 3) middle-range, and 4)

operational. The MTS is an early contribution resulting from this research.

Table 10 The four types of theory in a Multi-Theory Scaffold (MTS)

Type of theory Analytical value to the scaffold

Meta A way of thinking about other theories; a high level of abstraction; emphasizes ideas and concepts; includes grand and general theories

Critical Challenges paradigms and knowledge; sensitises researcher to a particular perspective (for example, emancipation of women, an axiological consideration); heightens researcher reflexivity

Middle range

Links abstract and pragmatic outcomes (meta or theoretical) to operationalised empirical research

Operational Specifies concepts and relationships as measurable variables

Including the four different types of theory (as proposed in Table 10) in the

theoretical scaffold can offer different perspectives to consider during the research

process. Theorisation may be stimulated by embracing both similarities and paradoxes in

theories. As Poole and van de Ven (1989) have recommended, recognising explanation at

different levels and taking advantage of theoretical tensions is useful for theory-building.

Just as complementary mixed methods offer more pieces of the puzzle, complementary

theories offer a more complete view of the picture.

...because theories may not be applicable as grand concepts explaining all incidents of phenomena, there is also room for simultaneously using more theories to encourage imaginative thinking, even if such theories have conflicting views on the relevance of particular concepts for understanding the issues studied as well as on how such concepts are interrelated. (Anderson & Kragh, 2009, p. 2)

A general discussion of the strengths and weaknesses of each of the four types of

theory in the MTS now follows. Section 3.2.4 furthers this discussion by illustrating how

different types of Human Agency theory within a MTS may be of value in understanding

the phenomenon under study.

1) META-THEORY

Meta-theory is a “theory about theories” (Shoib et al., 2006). Gregor (2006)

suggests that meta-theory, from a methodological perspective, is “a way of thinking

about other theories, possibly across disciplines (p. 616); it may serve as “a theory for

understanding” or a “sensitising device” (Gregor, 2006, p. 264). The strength of

including a meta-theory in a scaffold is that meta-theories offer an overarching

perspective of a study. General and grand theories are similar in the sense that they

provide broad explanation (Bacharach, 1989).

The limitations of using a meta-theory is that, like grand and general theory,

abstract theories often involve unique concepts and terminology not “easily used in

analytical theorizing” (Giddens, 1984, p. 162). Mills (1959) suggests grand theories are

so highly abstract28F

31 that they are of little or no use in research (Silverman, 2009).

However, Layder’s AT (1998) contends that there is room in theory use and development

for “a co-operative two-way borrowing” between general theory and empirical research

(p. 15). The limitation is not that such theory is “disengaged from the ‘real’ empirical

world”; rather, because of its high level of abstraction, its relation to the real world is

often “oblique and indirect” (Layder, 1998, p. 94).

2) CRITICAL THEORY

Critical theories, most often underpinned by a critical paradigm, include post-

modern and poststructuralist theory. However, the term ‘critical’ must be considered

(Danermark et al., 2002, pp. 200-201) as its meaning can vary. The term ‘critical’ is

often subject to “various interpretations” (Howcroft and Trauth, 2005, p.1) and

disciplinary connotations (Mingers, 2000).32 A critical perspective framework has been

summarised as: concerned with questioning assumptions; committed to emancipation;

taking a social perspective as opposed to an individualised one; and, sensitised to power

relations (Burgoyne and Reynold, 1997, cited in Wilson, 2004, p. 83).The inclusion of a

critical theory in a scaffold strengthens axiological considerations, which include “ethical

and moral questions” (Gregor, 2006, p. 612). For example, feminist critical theory

sensitises the researcher to the contingent conditions that women face and raises 31 Mills (1959) suggests grand theorists are “drunk on syntax” (p. 34). 32 Critical theorists are often informed by the Frankfurt school, for example, Adorno, Marcuse and Habermas (Hammersley, 1995), and are exemplified by Foucault and critical feminist theorists, including Wajcman (1991).

epistemological considerations such as the value of “seeking knowledge from below”

(Calás & Smircich, 2009, p. 265). In the IT research context, Trauth and Howcroft

(2006) propose that the objective of critical research about gender and IT is to investigate

why gender inequality exists, and to include central themes of emancipation, critique of

tradition, critique of technological determinism, and researcher reflexivity. A further

strength of critical research is that it can shift the attention from the individual to the

broader context of a phenomenon, thus including political, social, and economic

considerations (Doolin, 1998).

However, there are several limitations of solely utilising critical theory to

scaffold research. Critical theories may be essentialist towards gender if they emphasise

divisions (Adam et al., 2004) which may “valorise fixed gender identities” (Halford,

Savage, & Witz, 1997, p. 11). Further limitations involve a limited explanatory power

due to a focus on the ideological nature of participation, rather than on its empirical

outcomes (Trauth & Howcroft, 2006). Hence, critical theory may furnish “surface”

investigations, which then become the explanation (Trauth & Howcroft, 2006, p.5); this

may lead to discourse rather than explanation (Giddens & Turner, 1988, p. 160). Critical

theories may also focus on what is wrong with the world rather than what is right

(Walsham, 2006a), thus potentially recognising only the barriers and constraints on,

rather than the support of, women’s participation.

3) MIDDLE -RANGE THEORY

What is described as ‘middle-range theory’ (but somewhat differently by

Giddens, 1988, p. 82, & Merton, 1967) moves away from the greater level of abstraction

of meta or general theory because of its emphasis on “empirical data” and “empiricist

epistemology” (Layder, 1998, p. 138). Middle-range theory is thus strongly associated

with applied research at the “intersection of practice and research” (Smith and Liehr,

2008, Preface). The strength of including a middle-range theory in a scaffold is that these

theories are “concrete enough” to enable tests against empirical data (Danermark et al.,

2002, p. 118), and offer a “mid-way between the minor working hypotheses of everyday

life and the ‘grand’ theories” (Layder, 1998, p. 16). Thus, a middle-range theory can

provide an important link between the meta and operationalised theory in the scaffold.

However, there are also limitations in drawing solely on a middle-range theory.

Meta-theory is seen as being difficult to implement within empirical research because of

its high level of theoretical abstraction; by contrast, the limitation of middle-range theory

is its strong emphasis on empirical insights. Giddens (1988) points out that because

middle-range theories are limited to “a determinate range of conceptual issues” (p. 283),

any emerging “empirical generalizations” may still require some abstract formulation to

explain them (p. 164). There remains the need to consolidate them into a “progressively

more general conceptual scheme” (Giddens, 1988, p. 283). Consequently, the use of a

middle-range theory as a scaffold may well require another theory, such as an

overarching meta-theory, as a complement.

4) OPERATIONALISED THEORY

Operationalised theories comprise clearly defined constructs, propose measurable

concepts, and focus on a limited set of measurable variables when using objective data to

predict or confirm the relationship among factors and to test hypotheses or theories. 30F

The strength of operationalised theories, as a scaffold, is that they draw attention to

specific theoretical concepts and suppositions that can be tested in situated contexts.

There are, however, limitations in drawing on an operationalised theory. Meta- or

grand theory suffers from being a monolithic conceptual framework–from being

“nothing but theory”; operationalised theories, by contrast, are criticised for providing a

too-narrow abstracted empiricism, providing “nothing but statistics” (Mir & Mir, 2002,

p. 112). In focusing on a particular set of variables, operationalised theories may lead to

the researcher losing sight of the wider social context. As Giddens (1988) proposes,

when one starts with particulars, one rarely rises above them (p. 167). In response to

such criticisms, operationalised theories have drawn on other theories to broaden their

capacity for explanation. For example, Social Cognitive Career Theory (SCCT) as an

operational theory draws on Social Cognitive Theory (SCT), which is a middle-range

theory. The fact that operationalised theories can draw on other theories further

strengthens the argument for the use of multiple theories in the MTS.

PARADIGMATIC ASSOCIATIONS OF THEORIES IN THE MTS

The inclusion of different theories in a MTS may raise concerns of paradigmatic

commensurability of these theories. The stance taken is that the paradigmatic position of

a theory should not constrain its use in the MTS. Such a stance recognises the recent

“quiet methodological revolution” that has led to a “blurring” in paradigmatic positions

(Denzin & Lincoln, 2008, p. vii), and some “commensurability between paradigms”

(Guba & Lincoln, 2005, p. 117). By taking an “integrated paradigmatic” approach,

33 However, it should be noted that several methodologists contend that qualitative data can be used to test hypotheses (s Patton, 1990; Tashakkori & Teddlie, 2003).

theories are not viewed as dualistic or incommensurable, but rather as complementary

(Giddens & Turner, 1988, p. 135).

Nevertheless, the use of different theories in the MTS must recognise the

differences between paradigms because, as Layder (1998) notes, the use of theories can

have “antithetical” or “incompatible premises” underlying ontological and

epistemological assumptions (p. 135). However, juxtaposing theories from differing

paradigmatic stances also has benefits (Anderson & Kragh, 2009). A “multiparadigm

approach to theory building can generate more complete knowledge than any single

paradigmatic perspective” (Gioia & Pitre, 1990, p. 599). However, the stance taken in

this investigation is not multi-paradigmatic as advocated by Mingers (1997), or an

‘anything goes approach’ (Feyerband, 1999). Rather, an acknowledgment of paradigms

enables the researcher to embrace the tensions and paradoxes among theories. Embracing

paradoxes in a constructive way can lead to the development of new insights (Lewis &

Grimes, 1999). Employing a strongly pluralistic approach when choosing theories as a

scaffold, may strengthen the research approach.

3.2.3 HUMAN AGENCY THEORY

The previous section presented an argument for the use of multiple existing

theories as a scaffold, and proposed the MTS. This section introduces the theoretical

perspective of Human Agency as a way of understanding the problem of women’s

participation in the DCI. A MTS comprised of four Human Agency theories is proposed

in Section 3.2.4. This Human Agency Multi-Theory Scaffold (HAMTS)forms Framework

2, the second framework to be used for analysis of the empirical data.

A definition of agency finds it is “slippery” (Hitlin & Elder, 2007, p.170), and the

term maintains “elusive, albeit resonant, vagueness” (Emibayer & Mische, 1998, p. 262).

In most definitions, agency refers to an individual’s capacity for free will or action,

where such human capacity is linked to structural aspects34. 31F A broad range of theorists

consider human agency in their work in some form, including Durkheim, Marx, Weber,

Goffman, Parson, Mead, and Merleu-Ponty. Theorists with a stronger focus on human

agency include Giddens (1979, 1984) and Bandura (2001, 2002a).

In the broader sociological domain, gender-related research has recognised the

importance of human agency, where women’s emancipation involves their “individual

will and agency” (Connell, 1987, p. 50). Similarly, the literature regarding social

34 The term ‘agency theory’ is not used in the same way as it is in the economics literature (Eisenhardt, 1989b).

exclusion recognises the importance of women’s agentic capability (Sen, 2000; Zheng &

Walsham, 2008). However, agency theories have not been explicitly used to explore

women’s participation in the DCI. They may offer a new and relevant approach to the

research problem, particularly as they can address several of the methodological

concerns raised earlier in Section 2.3. This is because agency theory fosters an

exploratory approach, multilevel analysis, and is non-essentialist towards women.

The review of literature suggested the need to recognise the influence of the

interaction between the environment and person (See Section 3.1.3). Agency theories

recognise a dynamic rather than fixed model of interaction between these two entities.

For example, Giddens’ theory neither starts from, nor privileges, the subject or society

(Walsham, 1993). Similarly, SCT recognises that the dynamic “emergent interactive

agency” (Bandura, 2001, p.4) will differ based on the individual, the particular behaviour

being examined, and the specific situation in which the behaviour occurs (Bandura,

1989a; 1989b). This recognition of the dynamic nature of interaction provides a suitable

theoretical approach for exploratory, empirical research.

The review of literature suggested influences manifest over different levels, such

as macro and micro (See Section 2.5). Agency theories encourage multi-level analysis, as

they consider different levels of influences, including society and the individual. All

agency theories are concerned with the reciprocal relationship or interaction between the

environment (structure) and the individual (agent), even if different theories emphasise

different aspects such as temporality (Emirbayer and Mische, 1998), or the role of the

individual’s emotions (Archer, 2004). A strength of Giddens’ ST is that it attempts to

reconcile accounts of human agency and social structure in a way that bridges the

traditional micro-macro divide, where “studying social interactions in everyday life sheds

light on larger social systems and institutions” (Giddens, 1993, pp. 90-91). Indeed,

structure and agency may supersede macro and micro levels. As Ritzer and Smart (2001)

suggest, what was once “discussed under the rubric of the macro/micro problem is now

debated as the problem of structure and agency” (p. 350). Layder (2006) further refines

the terminology to suggest that micro-macro indicates the “level of analysis whereas the

agency-structure perspective can refer to large and small scale features of social life” (p.

Other theoretical approaches that offer a perspective on the interaction of the

environment and person face limitations in their potential to understand both macro and

micro aspects. For example, the limitation of drawing on theories arising from a

Symbolic Interactionism (SI)32F

35 perspective is that their focus emphasises micro-level

interaction, to the possible detriment of understanding the macro aspects. Layder (2006)

suggests it not so much that the focus of SI is on the ‘minutiae’ of interaction, but rather

that there is a lack of linkages between the face-to-face interactions with the wider

institutional contexts, and that SI may dismiss the “less readily observable” aspects of

interaction (Layder, 2006, p. 88). The exploratory research approach adopted in this

investigation would ideally, initially, remain open to a wide range of influences.

The review of literature suggested that women’s ‘gendered’ experiences need to

be understood in a non-essentialist manner (See Section 2.5). Agency theories offer a

way to transcend essentialist gender-based categories by offering a perspective that views

participation as an outcome of both individual agency and gender. Gender is not a fixed

category, but rather, a way of recognising how individuals are positioned within

interaction systems (Giddens, 1984). As Giddens (1993) suggests, simply labelling a

person or identifying a “set of characteristics” does not “indicate the process of

interaction” (p. 128). Similarly, Bandura (2002b) notes that “there is a substantial

difference between theorizing based on categorical cultural trait ascriptions and process

analyses” (p. 276).

Even though gender is not the central focus of agency theories, several remain

sensitive to gender by recognising that women face certain challenges. Giddens’ ST is

sympathetic to feminist themes (Mackrell & Nielsen, 2007). Connell (1987) goes so far

as to suggest that Giddens’ theory comes closest “to the requirements of a theory of

gender” of “all current frameworks for social theory” (p. 94). Giddens (1993) has

recognised that the obstacles women face in their careers include employers who

discriminate against women, as they believe women will leave the workforce to have

babies (p. 245). Similarly, Bandura (2002a) recognises that “socialization practices”

“undermine their [women’s and minorities’] efficacy for technological and scientific

careers” (p. 6). This is particularly relevant to this investigation as technology features

strongly in the DCI context.

35 Symbolic Interactionism includes the original work of Mead, Cooley and Dewey, and later theorists such as Blumer.

3.2.4 A HUMAN AGENCYMULTI -THEORYSCAFFOLD (HAMTS) Section 3.2.2 established that different types of theory in a scaffold bring a

different perspective or level of abstraction to exploring and understanding a

phenomenon. This section details the relevance of four agency theories, which will be

used to analyse empirical data. As Table 11 illustrates, the four agency theories align to

the four types of theories in the Multi-Theory Scaffold: meta, critical, middle-range, and

operationalized (as detailed in Section 3.2.2). Each of the theories in Table 11 provides

concepts that can foster an understanding of the environment and person influences and,

importantly, the interaction between the two, to reveal and help understand the influences

on women’s participation in the DCI. The choice of concepts is initially open, and further

refined during analysis as their relevance becomes apparent.

Table 11 Framework 2: the HAMTScomprised of four Human Agency-related theories

Type of theory

Specific theory Domain Key value to the current case

study Specific concepts that may be useful

Meta Structuration Theory

ST(Giddens, 1984)

Human agency

Provides overarching framework; considers

conditions (for example, structures/environment)

Modalities (norms, power, sanctions)

Critical

Social Theory of Gender STG (Connell, 1987)

Gender identity

Heightens researcher’s sensitivity to axiological

considerations, such as how a particular identity (for

example, gender) may experience conditions

differently to other identities

Middle range

Social Cognitive

Theory SCT (Bandura,

1989b; 1997, 2001)

Human agency,

Cognitive Psychology

Emphasises cognitive mechanisms an agent may

use when interacting with their environment (such as self-

efficacy)

Self-efficacy, disposition, scaffolding, feedback

Operational

Social Cognitive

Career Theory SCCT (Lent, Brown and

Hackett, 1994)

Human agency, Career theory

Operationalised SCT concepts (such as self-

efficacy)

Self-efficacy

As Figure 5 illustrates, the four agency theories form the HAMTS (Human

Agency MTS) form the second framework (Framework 2) used to analyse the empirical

data. Although all four theories are related to ‘agency’, they each offer a different but

complementary analytical value in the overall conceptualisation and explanation of the

phenomenon. Where a middle range theory–such as Bandura’s SCT–specifies constructs

regarding the role of an individual’s cognitive mechanisms (Bandura, 2001), a meta-

theory–such as Giddens’ ST (1984)–offers a different level of abstraction by providing an

overarching abstract concept, such as Structuration. The value to the research approach

of each of the four theories is now discussed.

Figure 5: The HAMTS theoretical framework (Framework 2)

1) META-THEORY : STRUCTURATIONTHEORY (ST)

Giddens’ Structuration Theory (ST) (1979, 1984) is one of the most widely

known agency theories. ST offers a highly versatile (Fayolle, 2011) theoretical approach

to understanding the everyday interactions between society (which involves the

structures and systems of the environment) and the stratified agent (involving the

thoughts and actions of the agent) (Giddens, 1988, p. 174).33F

36 Giddens’ ST encompasses a

range of concepts (Shoib et al., 2006). Concepts such as ‘duality’ have been used to help

explain how norms may influence women’s participation in IT (Beekhuyzen, Nielsen &

von Hellens, 2003), and how mentoring provides a way to challenge structures of

signification (von Hellens, Nielson, & Beekhuyzen, 2004). ST may offer a similar value

in understanding the phenomenon of women’s participation in the DCI.

Giddens’ work is relevant to research that focuses on the interaction between the

environment and person. Giddens’ stratification model of action (Giddens, 1979, 1984)

explains interaction or “interplay” (Giddens, 1979, p. 82) between society and the

individual. The model presents a set of three ‘modalities’ as a link or bridge between the

structural dimensions and the complementary dimensions that agents draw on when they

‘act’ in situated interaction. Modalities link the individual’s capabilities of

communication, power and morality to the structural dimensions of signification

(semantic rules), domination (distribution of resources), and legitimation (moral rules);

however, any distinction between the modalities is analytical rather than substantive

(Giddens, 1979, p. 55).

A second model, the stratification model of the agent (Giddens, 1984, p. 5)

recognises the agent’s own awareness where, through reflexive monitoring of action,

36 These are crudely akin to the categories of ‘society’ and the ‘individual’ (Giddens, 1984, p. 162).

HA-MTS

Analysis

Empirical data

“agents can explain, both to themselves and to others, why they act as they do by giving

reasons for their action” (p. 151). Giddens distinguishes between three levels of

consciousness which form a continuum between the unreflective and reflective

dimensions of action: the unconscious, practical consciousness (tacit, taken-for-granted

knowledge), and discursive consciousness (knowledge/reasons that can be verbally

articulated) (Giddens, 1979, p. 56). Thus, ST is compatible with a research approach that

aims to elicit women’s responses and capture participants’ reflexive knowledge.

It is proposed that Giddens’ ST can serve as a meta-theory in the MTS employed

in Framework 2. As a meta-theory, ST offers “a very high level of abstraction” of

concepts (Gregor, 2006, p. 616). Indeed, ST has faced criticisms regarding its abstract

nature and the fact that it provides little real practical guidance (Dobson, 2001b, p. 208).

This may explain, in part, comments made by Ramsey and McCorduck (2005) who

assert that although ST is a “promising” theory with which to understand the problem of

the lower rates of participation of women in the IT sector, it is “not mature enough to

build a program of action upon” (p.20). Although ST, as a meta- theory, may be difficult

to apply in an empirical study (Hatch, 1997; Pozzebon & Pinsonneault, 2005), several

researchers do offer insight into its application in such studies (Jones, 1999, p. 113;

Shoib et al., 2006; Rose, 1998).

Such criticisms strengthen the argument that complementary theories may

compensate for the perceived deficiencies of ST (Rose, 1998). It would not be judicious

to base an exploratory case study solely on this theory. However, aspects of the theory

may be useful. As noted earlier, when using theory as scaffold, select concepts may be

chosen. Indeed, Layder (2006) specifically suggests that “structuration theory” can be

drawn on in “small bits” or “whole chunks”, as a researcher considers them useful in

illuminating analysis (p. 156).

2) CRITICAL THEORY : SOCIAL THEORY OF GENDER (STG)

Connell’s (1987) Social Theory of Gender (STG) is a social structural theory

concerned with gender relationships. Although not explicitly an agency theory, there is

significant congruency between Connell’s STGand Giddens' ST. Connell’s (2002)

contemporary model describes four main, analytically separate, conceptual dimensions of

gender–production, power, emotional, and symbolic relations–all concepts presented in

Giddens' theory to some extent.34F

37 This may explain why Connell (1987) highlights that

37 Connell's later model was modified from the original theory proposed in 1987.

even though no theory of gender has been “couched in these terms”, Giddens offers the

“outlines of a solution” to the gender problem (p.62).

It is proposed that Connell’s (1987) STG may serves as a critical theory in the

MTS employed in Framework 2. As a critical theory, Connell’s (1987) theory heightens

the researcher's reflexivity and awareness of axiological concerns regarding

emancipatory aspects of women’s participation. In particular, STG draws attention to the

role of power, an influence identified in previous research in the ICT domain by Trauth

and Howcroft (2006). Additionally, as both Connell and Giddens draw attention to the

concept of power, STG may provide theoretical triangulation which, as Section 4.4.1

discusses, strengthens theoretical insights.

3) MIDDLE RANGE THEORY : SOCIAL COGNITIVE THEORY (SCT)

Bandura’s (1986) Social Cognitive Theory (SCT) is a Human Agency theory that

emphasises the role of a person’s cognitive mechanisms. SCT is also often linked to

learning theories and the psychology domain (Bandura, 1989b; 2001). SCT emphasises

emergent and dynamic interactions resulting from “the dynamic interplay” of three

groups of reciprocally determinate influences (Bandura, 1986). Two of these three SCT

categories are explicitly congruent with Framework 1 (the analytical framework): the

Environment (E), which includes environmental events; and the Person (P), which

includes personal factors (cognitive, affective, and biological). The third category in the

SCT model includes behavioural patterns (Bandura, 1986, 1999a). Although SCT

encompasses several other concepts–including temporal dynamics, fortuitous

determinants, and motivation processes–the key focus is on the cognitive mechanisms of

the individual, including: motivation and reflexivity (Bandura, 2006); self-efficacy

(Bandura, 1997); setting goals (or ‘intentionality’); envisaging outcomes (or

‘forethought’); the ability to react and adapt (or ‘self-reactiveness’) (Donovan, 2001).

It is proposed that Bandura’s SCT (1989b, 1997, 2001) may serve as a middle

range theory in the MTS employed in Framework 2. As a middle-range theory, SCT links

empirical and theoretical insights. For example, it encompasses concepts from both

Giddens’ ST (abstract concepts, such as cultural norms) and Lent et al.’s (1994) SCCT’s

operationalised variables (such as self-efficacy). SCT appears to be a useful theoretical

perspective; however, few previous studies have explicitly drawn on SCT.

Table 12 provides examples of how several concepts from SCT can help explain

previous research regarding women’s participation in the ICT industry. For example,

Adya and Kaiser (2005) posit “self-efficacy” (pp. 234, 238, 249, 248) as an important

aspect of girls’ participation in ICT. However, they position self-efficacy as an aspect of

Trauth’s IDT (Trauth et al., 2004), rather than Bandura’s SCT– even though IDT does not

explicitly identify self-efficacy as a factor. Similarly, Webb and Young’s (2005)

findings from a study investigating women’s participation in the Tasmanian ICT industry

propose a set of themes, several of which, such as skills and serendipity, resonate with

SCT. Other studies related to the phenomenon of women’s participation have drawn on

SCT directly. These include those concerned with gender related studies (Bussey &

Bandura, 1999), with media (Bandura, 2001), and technology (Bandura, 2002).

Therefore, SCT, as a middle-range theory, appears to offer a relevant scaffold for the

phenomenon under study.

Table 12 SCT concepts (Bandura, 1997) evident in previous research SCT SCT explanation As evident in previous studies of women’s

participation in the ICT industries

The environment provides models; observational learning occurs when a person watches the actions of another person

Positive role models have been cited as being a major factor in young women‘s career decision-making (Adya & Kaiser, 2005).

Reinforcements: Responses to a person’s behaviour that increase or decrease the likelihood of reoccurrence may promote self-initiated rewards and incentives

Trauth (2002, p.15) speaks of the importance of “selective reinforcement” of messages from society.

Behavioural capability: Knowledge and skill to perform a given behaviour; mastery learning through skills training

Beekhuyzen et al. (2003) identified that it is not a lack of technical skills or ability (or perceived skill), but rather behavioural attributes (perceived confidence or self-efficacy) that hinder participation.

Expectations: Anticipatory outcomes of a behaviour

“Social and cultural biases” influence “both the internal view that women have of themselves (self expectations) and the external view of women (stereotyping, for example) that is held by society in general” (Ahuja, 2002, p. 22).

Expectancies: The values that the person places on a given outcome; incentives

von Hellens, Nielson, and Beekhuyzen (2003) suggest that the way women talk about their experiences reinforces the stereotypes within the IT industry.

Self-control: Personal regulation of goal-directed behaviour or performance

Moore, Griffiths and Richardson’s (2005a) survey identified the need for “‘determination’,’ logical’, ‘persistence’” (p. 18).

Self-efficacy: The person’s confidence in performing a particular behaviour

Trauth (2002, p.109) reported that participants talk about ‘inner strength and self-confidence and self-esteem’. Moore et al’s. (2005a) survey of women working in IT noted that the women identified the need for “confidence and self belief” (p.18).

Emotional coping responses: Strategies or tactics that are used by a person to deal with emotional stimuli

Trauth (2002, p.110) identified that women’s reactions to barriers varied significantly. Webb and Young (2005) mentioned that other personality traits such as ‘adaptability’ may be key factors.

4) OPERATIONALISED THEORY : SOCIAL COGNITIVE CAREERTHEORY (SCCT)

Social Cognitive Career Theory (SCCT) 35F

38 considers the individual, their

context, their development and their interaction (Patton, Bartrum & Creed, 2004) in

relation to their interest, career choice, and performance, to explain the development of

their career interests and decisions. SCCT draws attention to how an individual can

exercise personal agency. It focuses on specific mechanisms that shape interests and

choices related to entry into, and persistence within, a profession. These mechanisms

include self-efficacy beliefs, outcome expectations and goal representations (Lent,

Brown, & Hackett, 1994, p. 83). SCCT has been empirically supported with a number of

studies, including studies of women and ethnic minorities in math, science and

engineering (STEM) (Lent, Brown, Schmidt, Brenner, Lyons & Treistman, 2003)36F

39 (Betz

and Hackett, 1997) and Information Technology (Smith, 2002).

It is proposed that Lent et al.’s (1994) SCCT may serve as an operationalised

theory in the MTS employed in Framework 2. Like the other theories in the HAMTS,

SCCTpositions participation as a result of “person-situation interaction” (Lent et al.,

1994, p.82). Person inputs include gender, and the situation or context includes

environmental variables such as socioeconomic status. As an operationalised theory,

SCCT provides a focus on specific constructs identified in the other theories in the

HAMTS.For example, it is informed by Bandura’s (1986; 1989b) SCT and,

consequently, posits self-efficacy as an important mechanism. SCCT, however, offers the

researcher an alternative perspective to SCT with respect to the causal pathways of

certain variables by proposing a greater direct influence of environmental factors (Lent et

al., 2003, p. 460). Furthermore, as several SCCT variables, such as contextual supports,

are reported as understudied (Lent et al., 2003, p. 459), these may be further considered

in the current case study if and when the empirical data draws attention to them.

PARADIGMATIC ASSOCIATIONS OF AGENCY THEORIES

As proposed earlier, recognising the paradigms underpinning theories can

strengthen the research approach as differences or tensions between the different theories

provide points of analytical stimulation. Identifying paradigmatic associations is,

however, somewhat subjective on the researcher’s behalf, particularly when a

paradigmatic stance is not articulated by the originators of the theories. For example, a

range of scholars has interpreted the ontological stance and consequent epistemological 38 Although the term ‘career’ is part of the SCCT title, academic interest, choice and performance are also part of the theory. 39 STEM: Science, Technology, Engineering and Mathematics

approach of Giddens’ ST differently. ST is seen as both subjectivist and objectivist;

Giddens is identified as both “rooted in a subjective ontology” and in “an interpretive

epistemology” (Shoib et al., 2006, p.142); he is also said to be “vehemently opposed” to

objectivism, and thus privileging “agents' views at the expense of other perspectives”

(Layder, 1998, p.141). 37F

40 In contrast to this, Giddens has also been labelled a structuralist

(Parker, 2000), accused of operating at a “distance” from the hermeneutic task and

subjectivism (Calhoun, Gerteis, Moody, Pfaff & Virk, 2002, p. 226). This apparent

contradiction may also be a result of the strongly pluralistic nature of Giddens’ theory.

3.2.5 SUMMARY

This section has responded to the general call in ICT research for an increased

awareness of the role of theory (Weber, 2003a, 2003b; Gregor, 2006). In proposing that

a theoretical framework may be beneficial for data analysis, methodological concerns

regarding the use of theory have been central. Building on the work of Layder and

Walsham regarding using theory as a scaffold, Section 3.2.2 argued for the value of

multiple theories in a scaffold and proposed the Multi-Theory Scaffold (MTS). A MTS is

an analytical tool that asks the researcher to consider the use of four different types of

theories in a scaffold; 1) meta, 2) critical, 3) middle-range and 4) operationalized. The

argument being that each different type of theory provides a different level of

abstraction. The Multi-Theory Scaffold (MTS) is an early methodological contribution.

Section 3.2.3 presented an argument for the use of human agency as a specific

theoretical perspective with which to understand influences on women’s participation.

Accordingly, Section 3.2.4 outlined the value of four key agency-related theories. The

meta-theory of Giddens’ Structuration Theory (ST)(1984), the critical theory of

Connell’s Social Gender Theory (SGT) (1987), middle-range theory of Bandura’s Social

Cognitive Theory (SCT) (1986, 1989b) and the opertionalised theory of Lent et al Social

Cognitive Career Theory (SCCT)(1994), as summarised in Table 11. These agency

theories form Framework 2: the Human Agency Multi-Theory Scaffold (HAMTS). Thus,

the HAMTS, is a specific instantiation of a Multi-Theory Scaffold (MTS). Framework 2,

provides several benefits in regards to identifying and understanding the influences on

women’s participation, including a non-essentialist approach towards gender and like

Framework 1 offers a focus on the interaction between the environment and person.

40 A website that lists resources for qualitative and interpretive research cites Giddens, with a reference to the Constitution of Society: Outline of the Theory of Structure http://www.qual.auckland.ac.nz/interp.aspx

This section has articulated the strengths of drawing on multiple agency theories

as a theoretical scaffold. However, the methodological approach must also consider

ontological underpinnings. Accordingly, the next section presents the argument for a

third framework with which to analyse empirical data, a framework stemming from the

ontology of Critical Realism.

3.3 FRAMEWORK 3: ONTOLOGICAL (CRITICAL REALISM)

To date, this chapter has outlined the benefit of employing two different

frameworks to analyse empirical data; 1) an analytical framework and 2) a theoretical

framework. However, research also requires a clear stance on the philosophical basis for

making both the epistemological and ontological choices (Wikgren, 2005). Such

ontological choices must be made explicit by a researcher (Archer, 1995; Walsham,

1995b). It is proposed that Critical Realism, in particular the ontological approach

proffered by Bhaskar (1979; 1986; 1989; 1994)38F, provides a suitable philosophical stance

from which to consider the phenomenon of women’s participation in the DCI41.

Accordingly, Critical Realism informs the development of the third framework used to

analyse empirical data.

3.3.1 A CRITICAL REALIST ONTOLOGY

It is challenging to provide an overview of Critical Realism as it provides

different levels of understanding, including: a philosophy of science, a realist social

theory, and an explanatory framework (Wikgren, 2005). It can also serve different

purposes such as: a philosophical underpinning (Mingers, 2000), a base paradigm

(Dobson, Myles, & Jackson, 2007), a philosophical “underlabourer” (Archer, 1998, p.

197)39F

42, and an “orientation” (Miles & Huberman, 1994, p. 4).

Although Lopez and Potter (2001) conclude that Critical Realism is not a

“homogenous field” (p. 311), but rather a “broad church” (p. 15), there is a general

agreement that Critical Realism stems from Bhaskar’s (1979; 1986; 1989, 1994)

transcendental realist philosophy. However, this research makes little direct reference to

Bhaskar’s (1978) work or explicitly draws on Bhaskar’s (1979) Transformational Model

of Social Action (TMSA). The primary reason why Bhaskar is not often directly

consulted lies in the difficulty in understanding some of his complex writings. Easton

(2009) suggests these may be better understood by drawing on Sayer (1992, 1997, 2000),

41 ‘Critical Realism’ is the label most researchers utilise; however, Bhaskar’s (1978) own label is “transcendental realism”. 42 Bhaskar describes transcentialism as an ‘underlabourer’.

whose “account of critical realist ontology is the most detailed and comprehensive” (p.

2). Rather than directly referring to Bhaskar, therefore, a range of different authors (such

as those presented in Table 13) form the main source of reference regarding Critical

Realism.

Table 13 Key Critical Realist authors informing the research approach Author As relevant to the current case study Bhaskar (1979, 1986, 1989, 1994)

Credited with proposing the underlying philosophy; focuses on the researcher’s attention on the real, actual, empirical stratified domains of reality

Archer (1995), Archer and Bhaskar (1998)

Builds on Bhaskar's transformational model and emphasises human reflexivity or ‘inner conversation’ as the central mechanism mediating the influence of objective social and cultural conditions; provides detailed consideration of Bhaskar and Giddens

Sayer (1992, 2000) Provides pragmatic insights into methodological aspects of Critical Realism

Danermark et al. (2002) Provides methodological guidance for Critical Realist research in the form of the ‘Six Stage Model of Explanation’

Layder (1993, 1998, 2006) Provides a broadly realist approach to theory use

A distinguishing feature of Critical Realism is the primacy of ontology (Lopez &

Potter, 2001, p. 75; Dobson, 1999, p. 265)40F

43, where ontology acts as “both gatekeeper

and bouncer for methodology” (Archer, 1995, p. 22). A Critical Realist ontology

acknowledges that objects of knowledge exist at two levels, intransitive and transitive

(Bhaskar, 1978, 1989; Archer, 1998). The “relatively enduring” world (independent of

human beings) is comprised of “intransitive” objects including structures, mechanisms

and processes, and events. Intransitive mechanisms generate the actual events that we

experience (Lawson, 1997; Mingers, 2004).44 Transitive objects consist of an agent’s

fallible knowledge of the world. The primary aim of Critical Realism is “to explain the

relationship between experiences, events and mechanisms” (Jeppson, 2005, p. 5), and the

emergent interaction between these, across a stratified reality (Easton, 2009).

There are ‘three domains of reality’: empirical, actual, and real (Bhaskar 1978;

Collier, 1994; Sayer, 1992). ‘Empirical’ refers to a person’s individual experiences that

can be observed and empirically measured; ‘actual’ refers to events themselves, whether

experienced or not; and ‘real’ refers to mechanisms, which can only be theorized about

and never fully understood (Collier, 1994). As Table 14 indicates, the ‘real’ domain

contains mechanisms, events, and experiences–the whole of reality; the ‘actual’ consists

43 Ontology is a branch of philosophy concerned with the nature and essence of things, and asks: ‘What kinds of things are there in the world?’ or ‘What is reality?’ Epistemology is the study of knowledge, and asks: ‘How can things be known?’ or ‘How can we know this reality?’ 44 This emphasis on ontology may explain why critical realism has also been more narrowly defined as the ontological aspect of a post-positivistic paradigm (Lincoln and Guba, 2000).

of events that do (or do not) occur, and the ‘empirical’ includes those events that are

observed or experienced (Mingers, 2004, p.93). Hence, understanding reality involves

examining the experiences of people, the events they experience, and the underlying

structures and mechanisms that make up the social world (Dobson et al., 2007).

Table 14 Ontological assumptions of the Critical Realist view Domainof real

Domain of actual

Domain of empirical

Experiences x x x The part of the real and the actual that is experienced by individuals

Events x x The domain of events

Mechanisms x The domain of causal tendencies

3.3.2 A CRITICAL REALIST FRAMEWORK

The third framework to be used to analyse empirical data is Framework 3, which

is based on the ontology of Critical Realism and is comprised of the three ‘domains of

reality’ that Bhaskar (1989) proposes: 1) empirical, 2) actual, and 3) real (as illustrated

by Figure 7).

Figure 6: Framework 3: ontological framework comprised of three concepts from Bhaskar’s Critical Realist ontology

Framework 3 can address several of the methodological considerations that the

review of literature in Chapter 2 introduced. The review of literature suggested that a

pluralistic paradigmatic approach would reveal a wider range of influences (See Section

2.5). Critical Realism balances different ontological perspectives by reconciling

incompatible ontology’s (Carlsson, 2006), and bridging objectivist or subjectivist

polarities (Olsen, 2008). Although Critical Realism is offered as an alternative to

positivism (Lopez & Potter, 2001) and postmodernism (Archer and Bhaskar, 1998, p. i),

and as an intermediate philosophical stance between realism and critical theory (Lincoln

Analysis

Empirical data

Empirical (Entities) Interpretation of participants’ experiences

Real (Mechanisms and processes)

Actual (Events and actions)

Interpretation of causes or tendencies, underlying mechanisms

Interpretation of patterns within empirical data

Ontological Framework (Critical Realism/Bhaskar’s three domains of reality)

& Guba, 2000), it does not encourage plural ontology’s; rather, it encourages

differentiated realities (Sayer, 2008).

Critical realism differs from other paradigms more often adopted in previous

gender and ICT research, such as social constructionism.42F

45 Iconic works, such as The

Social Construction of Reality (Berger & Luckmann, 1966), do not dispute that

interactions occur between the person and their environment, but ontologically believe

that they can only be understood from the shared knowledge and individual meanings

that people develop from their experiences of interactions. In contrast, for the Critical

Realist, social reality is not limited to what people know (Lopez & Potter, 2001, p. 28).

Although Varella and Harre (1996) emphasise methodological individualism, rather than

attribution of causal powers to social structures, the Critical Realist approach generally

aims to avoid a “myopic analytical focus on situated social interaction” to the “detriment

of the explanatory power of structure” (Reed, 1997, p. 25). For Critical Realism, the

subjective meanings of the person and objective structures both have a causal capacity

(Olsen, 2008; Archer, 1995). Thus, Critical Realism’s pluralistic paradigmatic approach

fosters subjective and objective insights into the phenomenon under study.

The review of literature suggested the need for a multi-level analysis, because

there are a diverse range of possible influences on women’s participation (See Section

2.5). Critical Realism fosters a multi-level analysis; for the most part, it does this by

recognising that reality manifests across its three domains–from the experiences of a

person through to the underpinning mechanisms. A Critical Realist perspective involves

moving beyond macro and micro distinctions; it does not concentrate solely on a single

level of investigation of the society, group or individual. Epistemologically, these levels

may include the biological, the psychological, the social, and the cultural (Wikgren,

2005). Furthermore, these levels must be considered from a “relational perspective”

(Dobson, 2002), thus acknowledging the interactive nature of society and individual

(Archer, 1995; Bhaskar, 1989, 1998; Collier, 1994). A perspective encompassing a

stratified reality is seen as providing a greater capacity for explanatory power than the

“flat ” ontology’s (Reed, 1997, p. 24)43F

The review of literature also suggested the need for a theoretical understanding of

women’s participation (See Section 2.5). Critical Realism’s emphasis on the need for an

45 Social constructionism is typically associated with Papert’s learning theory and suggests people learn through interaction; thus, it is described as ‘a sociological construct’. Social constructivism is typically associated with interpretive and critical epistemologies as there is an emphasis on how people create their own meaning; thus, it is described as ‘a psychological construct’. 46 Reed identifies these perspectives as ‘ethnomethodology’, ‘actor-network theory’ and ‘Foucauldian post-structuralism’.

explanation of the phenomenon through the identification of causal tendencies facilitates

the theorisation of “the nature of individuals, the nature of society” and “the nature of

their causal interaction” (Sayer, 2000, p. 140). Causal powers or causal liabilities exist

across a stratified reality, including human actions and social structures, where a “cause

is whatever is responsible for producing change” (Sayer, 2000, p. 94). Critical Realism

encourages both a description of the entities involved in the phenomenon–that is, the

“characteristics of the interaction of particular kinds of ‘things’”–and an explanation of

causal mechanisms underlying that interaction (Lopez & Potter, 2001, p. 11).

This key concern with causality and the identification of the causal mechanisms

in social phenomena is the “key” aspect of Critical Realism (Layder, 1993, p. 16).

However, the emphasis on generative mechanisms (Bhaskar, 1979, p. 170) is not “cause

and effect” in the positivist sense (Mutch, 1997). Rather, Critical Realism recognises that

“causal mechanisms do not always generate a particular sort of event”, but that there are

“tendencies to interact in certain ways” (Lopez & Potter, 2001, p. 11). Causality is not

for prediction; rather, it is a way to illuminate and explain what happens (Ackroyd &

Fleetwood, 2000, p. 15). It is from an understanding of this causation that one may build

theory (Sayer, 1992; Montona & Szmigin, 2004). Thus, Critical Realism’s focus on

underlying mechanisms and causal tendencies can move the research approach beyond

description as a sole outcome, towards explanation of the phenomenon.

The review of literature in Chapter 2 further suggested the need for an

exploratory research approach. Importantly, Critical Realism provides a clear ontological

approach congruent with exploratory research. Critical Realism does not require a

decision to be made about what entity or object is central in a study; it does not prescribe

which “rocks to look under” (Dobson et al., 2007, p.149). Epistemologically, there is

“no objective ‘primacy’ of one factor over others” (McLennan, 1989, p. 263), other than

those that have “primacy” as local variables “relative to a specific set of concerns”

(McLennan, 1989, p. 32). That is, the only influences that have primacy are those that

emerge through the situated interaction under study. Therefore, any explanation depends

upon identifying “powerful particulars” from the emergent data (McLennan, 1989, p. 3).

3.3.3 SUMMARY

As Section 3.3.1 introduced, the ontological stance of Critical Realism can offer

the underpinning paradigm for an investigation seeking to understand women’s

participation in the DCI. As Section 3.3.2 presented, Critical Realism provides three

specific concepts to be used in Framework 3 when analysing the empirical data; these are

the three ‘domains of reality’ proposed by Bhaskar: 1) empirical; 2) actual; and 3) real

(Bhaskar, 1978, 1989; Sayer, 1992, 2000). Like the two other frameworks, Critical

Realism encourages a consideration of the interaction between the environment and

person. Insights into the ‘empirical’ domain value the insights that participants have of

their experiences. An insight into the ‘actual’ domain, on the other hand, entails a greater

level of supposition from the researcher to identify events that participants experience

(be they aware or not aware of the events). Insights into the ‘empirical’ and ‘actual’

domain lead to identifying underlying causal mechanisms in the ‘real’ domain. The

identification of mechanisms is a key aim for Critical Realist research, and ultimately,

this research focuses on abstracting mechanisms.

3.4 CHAPTER SUMMARY

Chapter 2 established two research questions: ‘What are the influences on

women’s participation in the DCI?’ and ‘How might we understand these influences?’.

It was proposed that analysis of empirical data will reveal the influences, particularly if

the data includes rich descriptive insights from women working in the industry. Chapter

3 has proposed that influences may be both identified and understood by employing three

frameworks with which to analyse the empirical data: 1) analytical, 2) theoretical, and 3)

ontological (as summarised in Table 2).

Section 3.1 identified Framework 1, an analytical framework which emerged

from the review of academic and non-academic literature regarding women’s

participation in the DCI or related ICT industry. The literature suggests that influences

manifest across different levels, including: the Environment (E) at both macro (for

example, society) and micro (for example, organisational) levels; and the Person (P).

However, the interaction of these influences warrants greater attention. Accordingly,

there are three categories in Framework 1: 1) Environment (E), 2) Person (P), and 3) the

Interaction (I) between the two. Analysis of the empirical data using Framework 1 may

assist in identifying and understanding influences.

Section 3.2 identified Framework 2, a theoretical framework comprised of four

Human Agency theories, including Giddens’ (1984) Structuration Theory (ST), and

Bandura’s (1984) Social Cognitive Theory (SCT). Agency theories offer a suitable

theoretical perspective from which to understand the interaction between the person and

the environment. It was argued that the methodological approach of using multiple

theories as scaffold, was relevant for the problem at hand (See Section 3.2.2). An early

contribution emerged with the explication of the Multi-Theory Scaffold (MTS) in Section

3.2.4. The MTS provideds a tool to help guide the researcher’s choice of four theories to

scaffold research, where each type of theory provides a different level of abstraction. The

choice of the four Human Agency theories lead to the development of the Human Agency

Multi-Theory Scaffold (HAMTS) that forms Framework 2. Analysis of the empirical data

using Framework 2 will assist in identifying and understanding influences.

Section 3.3 identified Framework 3, a philosophical or ontological framework

based on Critical Realism. A Critical Realist perspective addresses several of the

methodological concerns raised in Chapter 2, such as the need for a pluralistic

paradigmatic approach and for multilevel insights into the phenomenon. Furthermore,

Critical Realism is compatible with a research approach investigating the environment,

the person, and the interaction of the two. Framework 3 entails analysis of empirical data

utilising Bhaskar’s (1986) ‘three domains of reality’: 1) empirical, 2) actual, and 3) real.

It is expected that analysis of the empirical data using these three concepts, underpinned

by a Critical Realist ontology, will assist in identifying and understanding influences.

As will be explained further in Chapter 4 these three frameworks emerged during

the research process, they were not developed prior to data collection but rather followed

progressively and after initial data analysis. The literature had guided the approach for

initial data collection; however, it soon became apparent that frameworks would help

both identify influences and offer a way to understand those influences in a holistic

manner. This chapter has described each of the frameworks individually; however, the

early synergies between the frameworks has been noted. For example, each of the three

recognise that there is an interaction between the environment and the individual.

In some ways, the proposed research approach appears complex: it employs three

frameworks involving multiple theories, and a pluralistic paradigmatic stance that

encourages both subjective and objective insights across multiple levels of analysis.

However, it was acknowledged at the start of this chapter that a complex issue such as

women’s participation may well require complex solutions. What follows from the

review is the need to specify a detailed research approach or plan for the conduct of the

research. Accordingly, the following chapter now details this research design.

Table 15 The three frameworks to address methodological concerns Methodological considerations

Framework 1 Analytical

Framework 2 Human Agency (Theoretical scaffold)

Framework 3 Critical Realism (Ontological)

1. Need for empirical description of influences from women in the industry

Little empirical research in industry context; previous research in ICT identifies need for participants’ perspectives

Epistemologically, agency theory sensitive to agent’s perspective; for example, ST recommends hermeneutic analysis

Empirical domain of reality values insights from person; values multiple subjective realities, but not limited to it

2. Need for multilevel and relationalinsights

Existing literature points to relevance of wide range of influences, including Environment and Person

Agency theories encompass wide range of entities, including those identified in previous literature

Considers a broad range of entities and relationships across a stratified reality; comprised of three domains of reality

Previous research suggests a diverse range of influences across different levels; for example, macro and micro

Acknowledges both structural and individual causal powers at macro and micro, environment and individual levels

Focus on both objective and subjective insights, including influences participants are not aware of

3. Non- essentialist approach to gender

Focus primarily on the individual agent, while remaining sensitive to gender

Need to explore interaction between influences

Although interaction seen as crucial, literature not clear on how to conceptualise interactions

Agency theorists such as Giddens and Bandura provide models of interaction; Interaction itself seen as emergent

Critical Realism provides ontological/ epistemological framework to explore interaction; Identification of causal tendencies in interaction

4. Pluralistic paradigmatic insights

Previous research emerges from different paradigmatic positions

Ontologically, agency theories such as ST bridge paradigms;MTSacknowledges the value of theories from different paradigmatic stances

Bridges paradigmatic positions; encourages investigation of subjective and objective insights

5. Theoretical explanation; not description only

Limited theoretical explanations of the phenomenon; theories positioned from different paradigmatic stances

Agency theory provides a baseline for theorisation; multiple agency theories may foster triangulation

Draws on both empirical data and theory to identify underlying causal tendencies or ‘mechanisms’; provides a form of explanation; uses different types of logic (for example, inductive/ deductive/ retroductive)

Chapter 4. Research Design ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 89 of 401

Chapter 4: Research Design

Chapter 2 introduced a number of methodological considerations for this

investigation, such as the need for empirical data and theoretical understanding of the

phenomenon of women’s participation in the DCI. These concerns included: 1)

empirical insights, 2) avoiding essentialism, 3) multi-level analysis, 4) pluralistic

paradigmatic insights, 5) theoretical understanding and 6) applied findings.

In response, Chapter 3 identified three frameworks to be employed in the analysis

of empirical data. As was noted at the end of chapter 3 these frameworks emerged during

the highly iterative research process. In qualitative study, “research design should be

reflexive process operating through every stage of a project” (Hammersley and Atkinson,

1995, p. 24).

This chapter revisits the concerns raised in chapter 2 and articulates the specific

details of the research design employed in the investigation, including: the use of case

study (Section 4.1), the pluralistic data collection methods (Section 4.2), the value of

Danermark et al.’s (2002) Six Stage Model of Explanatory Research to further guide the

data analysis and theorisation process (Section 4.3), the issues of research quality

(Section 4.4) and concludes with the concerns for theory building (Section 4.5).

As Section 4.1 details, case study provides a suitable strategy of enquiry for

several reasons, including its compatibility with Critical Realist ontology. It is vital to

explicate ontological assumptions because, as Morgan and Smircich (1980) noted, much

of the debate about research methods is “linked directly to assumptions about ontology,

epistemology, and human nature” (p.491). Critical Realism argues strongly for the

recognition of the relationship between philosophy and methodology (Dobson, 2001b,

2002): “methodology without ontology is blind” (Archer, 1995, p.28).

Section 4.2 discusses the sampling strategy and the pluralistic data collection,

where primary data includes interviews and secondary data includes industry documents.

The linkages to methodological concerns raised in chapter are revisited. For example,

chapter 2 identified the need for rich empirical data of women’s experiences. This

chapter recommends the method of semi-structured interviews to capture suitable data

from the participants.

Section 4.3 discusses how data analysis is further guided by Danermark et al.’s

(2002) Six Stage Model of Explanatory Research, an approach underpinned by Critical

Realism. The actual analysis techniques include several established qualitative

techniques such as pattern interpretation. Chapter 2 identified the need to investigate

influences across multiple levels, and to reveal influences that may be difficult to

observe. Thus, this chapter discusses the use of techniques such as hermeneutic analysis

to reveal these less readily observable influences.

Section 4.4 considers issues of research quality, such as researcher reflectivity

and ethics. As noted earlier, an intended outcome of the research is for a theoretical

understanding of the phenomenon of women's participation. Accordingly, the chapter

closes with considerations of theory building (See Section 4.5). The chapter notes at

several points that although the research process is reported in a linear manner, the actual

process is iterative. As Maxwell (2005) notes, often in qualitative research, “the activities

of collecting and analysing data, developing and modifying theory, elaborating or

refocusing the research questions…are usually all going on more or less simultaneously,

each influencing all of the others (p. 2).

4.1 CASE STUDY AS A STRATEGY OF ENQUIRY

There are five strengths of employing case study for the investigation.

1. Case study is highly congruent with Critical Realism. 2. Case study provides a suitable research approach for the research problem. 3. Case study provides an established strategy for managing data. 4. Case study can aim for both description and explanation (in the form of theory

development) of the phenomenon. 5. Case study is congruent with the use of existing theory.

1) Case study is a suitable strategy of enquiry for this investigation as it is highly

congruent with the underpinning ontological stance of Critical Realism. The case study

strategy is adaptable, offering a bridge across paradigms (Luck, Jackson, & Usher, 2006).

Case study can be used in paradigmatic approaches including positivist (Yin, 2003),

critical (Walsham, 1995b), and interpretive (Eisenhardt, 1989; Stake, 1995) research; it is

particularly compatible with Critical Realism (Tsoukas, 1989; Easton, 2001, 2005, 2009;

Yeung, 1997; Perry, 1998; Montano & Szmigin, 2004; Anderson & Kragh, 2009).

Although Critical Realism is a relatively new approach to ontological issues (Easton,

2009), it offers much in the way of a philosophical perspective. It has been applied across

a wide range of fields including geography (Yeung, 1997), organisational studies

(Ackroyd & Fleetwood, 2000; Reed, 2001), and sociology (Sayer, 1992). Of the few

empirical studies which apply a Critical Realist approach, the majority utilise case study.

This includes studies in information and communication technologies (Fox, 2009), South

African SMEs (Jeppson, 2005), and industrial marketing (Easton, 2009). In the industrial

network field, Easton (2000) suggested that Critical Realism offers a “philosophical

defence” for case study (p. 206).

From a Critical Realist perspective, case study can “be used to describe empirical

events and by its use of multiple data sources, trace out links over time, digging ever

deeper, and following through the actual to the real domains” (Easton, 2000, p. 211).

Critical Realism values the “standard features of case study” such as the “emphasis on

actor’s accounts, reliance on explicit theorizing and attention to sequencing and

dynamics of social processes over time” (Mills et al., 2009, p. 255). It is proposed that

case study is a suitable strategy with which to reveal the less visible influences on

However, Critical Realism provides less in the way of explicit methodological

guidance (Yeung, 1997; Carlsson, 2006). There are only a few examples of empirical

Critical Realist studies, particularly in the ICT domain (Dobson, 2001a, 2002; Mingers,

2003), although researchers such as Volkoff, Strong and Elmes (2007), and Wynn and

Williams (2008), have provided recent advances. Because there is “no such thing as a

method of Critical Realism” (Danermark et al., 2002, p. 73), researchers are left to work

towards identifying a method that can be readily applied for conducting empirical

Critical Realist studies (Raduescu & Vessey, 2009). However, as Yeung (1997)

suggested, a Bhaskerian Critical Realism is not so much “a philosophy in search of a

method” (p. 51); rather, in fact, it leaves the theoretical and methodological work to each

substantive domain. Thus, Critical Realism fosters openness to the choice of

methodological approach. Rather than having a commitment to any single form, Critical

Realism recognises the contribution that research methods from other paradigms can

make (Mingers, 1997, 2001), thus providing a starting point for the evaluation of already

established methods (Danermark et al., 2002, p. 73). Ideally, the choice of method in any

research is dependent on: topics and contexts (Layder, 1993); the purpose of the research

(Hammersley & Gomm, 2000); the nature of the object under study; and the “purpose

and expectation of the study” (Sayer, 1992, p. 4). Critical Realism, therefore, “endorses

or is compatible with a relatively wide range of research methods” (Sayer, 2000).

Qualitative techniques are compatible with case study (Stake, 2000, p. 435).

However, Critical Realism makes a distinction between ‘qualitative and quantitative’

methods, with Sayer preferring to describe them as “intensive” or “extensive”

(Danermark et al., 2002, p.2). The intensive/concrete involves theoretical and empirical

analysis (Sayer, 1992, p. 237), and mainly qualitative methods and analysis, to obtain in-

depth knowledge of specific phenomena (Jeppson, 2005). A case study approach offers a

research approach compatible with the underpinning ontology of Critical Realism.

2) Case study is a suitable research strategy for the research problem, being

particularly useful in newer, less well-developed research areas (Dobson, 2001b) and

exploratory research (Yin, 1994). In addition, Montano and Szmigin (2004, p. 363)

suggested that case study and Critical Realism are both suitable for understanding

complex social situations. As Chapter 2 identified there is very little empirical or

theoretical research into the phenomenon of women’s participation in the DCI.

3) Case study provides an established strategy of inquiry44F

47 (Yin, 1994; Denzin &

Lincoln, 2003). Thus, it can help counter the limited guidance in method proffered by

Critical Realism. However, case study itself does not have an established approach to

analysis (Miles & Huberman, 1994), and this makes analysis a difficult aspect of the case

study approach (Stake, 1995). Consequently, the implication for this investigation is that

established analysis approaches compatible with Critical Realism must be employed. As

is discussed in Section 4.3, Danermark et al.’s (2002) model of explanation can address

the analytical limitations of the case study approach.

4) Case study can aim for both description and explanation as outcomes.

Andersen and Gamdrup (1990) contended that there are five ‘ideal types’ of research

including: predictive, action orientated, explorative, descriptive, explanatory (p. 41); the

latter three were also identified by Yin (1993) as outcomes of case study. Dobson

(2001a) noted that when conducting a case study, “the concept of explanation suggests a

realist approach”, and description suggests more of an “interpretive approach” (p. 283).

Thus, case study can combine empirical, theoretical and ontological insights to deliver

descriptive and explanatory outcomes. However, as Neuman (1997) reminds us, the

distinctions between these outcomes blur “together in practice”, as explanatory research

builds “on exploratory and descriptive research” (p. 20). Like case study, a Critical

Realist approach encourages both a description of the key entities and an explanation. As

Table 16 illustrates this investigation will deliver explorative, descriptive, and

explanatory outcomes

Table 16 Case study outcomes resulting from the research conducted in this dissertation Outcome Yin’s(1993) definition Evident in dissertation Exploratory To explore what is happening;

data collection occurs before theories or specific research questions formulated

RQ 1 offers an initially broad research question. Framework 1 provides a broad, guiding analytical framework from literature on which to base further inductive analysis.

Descriptive To describe what is happening; theory can guide the collection of data.

Framework 2 encourages analysis of empirical data guided by the theoretical scaffold.

Explanatory To explain why it is happening; looks for cause-and-effect relationships

Framework 3 encourages analysis of empirical data by focuses on abstracting the causal tendencies.

47 Denzin and Lincoln use ‘inquiry’, rather than ‘enquiry’. This usage may reflect a difference between English and American spelling.

5) Case study is congruent with the use of existing theory. Theory can serve as the

‘design’ for the descriptive case study (Yin, 1993, p. 22), or as a scaffold (Walsham,

1995b, 2006). Eisenhardt (1989) argued that binding emerging theory with existing

literature strengthens the theory building from case study research. Hence, there is a

resonance between case study and the use of the theoretical scaffold (Framework 2)

outlined in detail in Section 3.2.2. Table 17 summarises several additional reasons

why case study is relevant to the current study.

Table 17 Relevant characteristics of a case study approach Characteristics of case study As relevant to the reported case study Point

Investigates a contemporary phenomenon within its real-life context (Yin, 1994)

A focus on participants’ experiences is congruent with Critical Realist ‘empirical’ domain of reality

Reveals a particular event in a real life context (Eisenhardt, 1989); events include individual lifecycles and organisational processes (Yin, 1994, p. 13)

Focuses on events participants experience; congruent with Critical Realist ‘actual’ domain of reality

Encourages multiple interpretations (Walsham, 1995b)

Critical Realism values subjective insights; however, is not limited to them

Makes use of multiple sources of evidence (Yin, 1994)

Critical Realism encourages triangulation 1

Use of rival (Yin, 1993) or complementary explanations (Walsham, 1995b); may not even entail any ‘field data’ (Yin, 1994)

Typology of theories (for example, the MTS )

Phenomenon is “looked at in depth” (Stake, 1995). 45F

48 Congruence with a focus on underpinning mechanisms

Suited to “new research areas or areas for which existing theory seems inadequate” (Eisenhardt, 1989, pp.548-549)

Need for theoretical understanding of a little- researched phenomenon and context

4.1.1 CASE STUDY CONTEXT, BOUNDARY AND UNIT OF ANALYSIS

As Yin (1994) proposed, a case study is “an empirical investigation that

investigates a contemporary phenomenon within its real life context” (p.14), and is useful

when the “boundaries and context are not clearly evident” (p.13). However, case study

design requires the identification of a clear boundary (Yin, 2004, p. 165), particularly

when the researcher seeks to understand a specific situation or context (Eisenhardt,

1989). This is so even if the boundaries between phenomenon and context are not readily

distinguishable (Yin, 1984, p. 23). Case study boundaries can be identified by sampling

parameters (Miles & Huberman, 1994, p. 38) such as settings, participants, time, space

(Creswell, 1998, p. 61), or process (Yin, 2003). For example, social organisational

settings may be considered as a boundary (Morgan & Smircich, 1980; Orlikowski &

Baroudi, 1991).

Specification of the context is important to the Critical Realist (Layder, 1993;

Danermark et al., 2002, p. 168) as it facilitates the identification of the characteristics of 48 Stake’s comments are in regards to intrinsic case study.

the entities involved in the phenomenon under study. In the broader literature, there are a

number of contending definitions of ‘context’ (Dervin, 1997). In studies of women’s

participation or under-representation in the ICT field, Trauth et al. (2005) proposed that

‘context’ can include geographic characteristics such as state, provincial or regional, or

national characteristics.

A case study approach is suitable for an investigation focusing on a specific

context, such as the DCI. However, the phenomenon of women’s participation extends

beyond the organisation or (even) industry context and wider society; thus, the

boundaries of the phenomenon of participation are not readily identified. As Table 18

summarises, the context of the case study includes the organisational context,

geographical, and temporal context.

Geographically, the sample population are participants working in Brisbane, the

capital of the state of Queensland in Australia. Brisbane may constitute what Pratt (2000)

identified as a ‘new media hub’, as the Queensland DCI employs more people in games

production than any other state (ABS, 2008a)46F.

49 Brisbane is also well represented with

multimedia production organisations. Thus, a city such as Brisbane reflects the various

dimensions of the Australian DCI (organisation size, type, and roles) within a relatively

contained geographic location. Because Brisbane provides a diverse range of DCI

organisations, it is seen as suitable location context for an exploratory case study; this

diversity may reveal different aspects of the phenomenon under study.

Table 18 Context of the case study Context As implemented in this case study Geographical Brisbane (the capital city of Queensland, a state in Australia)

Temporal 2007-2011 (data collection period) Setting Both games and multimedia production organisations Organisation type Start-up, multinational, public, private in the Australian DCI Participants Informants Women employed in the DCI in Interactive Content Creator roles Industry stakeholders

Case study also requires the identification of the unit of analysis (Yin, 1994). For

this investigation, the unit of analysis for the case study is ‘interaction’; this is reflected

in the inclusion of the Interaction (I) category in Framework 1, and the acknowledgement

of interaction as a central concern for agency theory (Framework 2) and Critical Realism

(Framework 3). As Figure 7 implies, any exploration of interaction must consider the

Person (P), and their Environment (E), as constitutive entities of Interaction (I).

However, as noted earlier in Chapter 2 and 3, the Critical Realist perspective encourages

49 Queensland was followed by Victoria and the Australian Capital Territory.

Primary Unit of analysis

multiple levels of analysis. The exploratory nature of this investigation entails a certain

openness to the data analysis.

Figure 7: Illustration of the context and ‘unit of analysis’ of the reported case study

4.2 DATA COLLECTION

This section discusses the study’s pluralistic and flexible approach to data

collection, which involves questionnaire, semi-structured interview and a wide range of

secondary data. Even though the study systematically used established tools and

processes such as a question bank, there remained an inductive approach to

understanding the data.

4.2.1 PARTICIPANT SAMPLING

Flyvbjerg (2006) noted that there are specific strategies for the selection of

participants when using the case study approach. In qualitative enquiry, these sampling

strategies include: variation, snowball (or chain), stratified, purposeful, and convenient

strategies. Sampling for variation is congruent with Critical Realism, as increasing the

range of data collected can increase the multiple realities that are uncovered (Mills et al.,

2009, p. 602). Sampling for variation is also a suitable approach for an exploratory case

study because any patterns that emerge may be of particular interest (Patton, 1990, p.

In the conduct of this research, the case study employed several sampling

strategies including maximum variation, purposeful, and opportunistic strategies (as

summarised in Table 19). The purposive sampling strategy limits participants (or

informants) to women employed in the occupational role of ‘Interactive Content

Creation’ class 267 (ABS, 2008b). A maximum variation sampling strategy captured a

diversity of occupational roles within the interactive Content Creator category and across

a range of organisation types and size (that is: start-up, multinational, private/commercial

and public/government). Similarly, stakeholders (such as both male and female

employers, employees, and industry representatives) are sampled to provide different

insights.

Context =DCI

Sample =Women in DCI

Environment (E)

Person (P)

Interaction (I) Wider context = society

Phenomenon of participation

Table 19 Examples of sampling strategies in the reported case study Strategy Reason adopted As implemented in the reported case study Purposive Informant with required

qualities Female interactive content creators working in the Australian DCI

Maximum variation

Greater range of insights suitable for an exploratory study

Informants employed in a range of roles (for example, design and programming) and organisations (for example, start-up and multinational)

Opportunistic Unplanned opportunities for data collection

Researcher responds to emerging data sources47F

Snowball Identification of participants who are rich data sources

Participants or industry contacts recommend informants

Further variation can be found in roles that are ‘embedded’, as opposed to ‘core’,

and in roles that are ‘public’ as opposed to ‘private’. In ‘core’ production, the creation of

digital content is the core business of the organisation. ‘Embedded’ production is the

development of digital content to support the organisation’s primary business, “for

example, web pages or advertising material for a law firm” (DCITA, 2005, p.6).

Although women may be employed in either type of role, it is assumed that core roles are

strongly linked with the DCI (in contrast to other industries) and should form the focus

of this research.

The research employed a suitable sample size for the method and strategy

employed. The aim was not for statistical representation, but rather for suitable data

analysis. Onwuegbuzie and Leech (2005a,b, 2007) provided several rationales for

systematically selecting sample size in qualitative research studies, the number

depending on the design approach. However, the actual number of people chosen to be

interviewed requires a “flexible response” (Layder, 1998, p. 70). For qualitative studies,

Morse (1994) recommended at least six (p.225) and Bertaux (1981) suggested fifteen as

the smallest acceptable sample (p.35). Creswell (1998) recommended 20-30 people for a

grounded theory design in order to “achieve detail in the theory” (p. 113) and 5-25 for

phenomenological designs (p. 64). Green and Thorogood (2009) stated that “the

experience of most qualitative researchers is that in interview studies little that is 'new'

comes out of transcripts after you have interviewed 20 or so people” (p. 120). Similarly,

case study method may focus on a small number of people “nested in their context and

studied in-depth” (Miles & Huberman, 1994, p. 27). Accordingly, the sample size of

twenty-one participants, 18 female DCI workers and 3 stakeholders, is suitable.

Although case study aims for analytical rather than statistical generalisation, the

sampled population may be considered from a quasi-statistical approach. Chapter 1

indicated that there was a 10.7% participation rate of women in the Australian games

industry in the 2006-2007 financial year. According to the ABS, 695 persons were

employed in games development in Queensland during that period. If it is then assumed

that 10.7 % of the 695 persons are female, this would equate to 74 women working in the

Queensland games industry. Therefore, it can be supposed that the sample of 9 women

from games production equates to over 10 % of the population of women working in the

Queensland context. 49F

51 Of course, this is not statistically sound; rather, it is a supposition

that aims to provide an indication of the sample as relative to the number of women

employed in the industry. A similar logic cannot be applied to the multimedia industry

context for the reasons outlined in Chapter 1 regarding the availability of accurate

employment figures.

4.2.2 DATA COLLECTION

The research design aimed to collect multiple forms of data, which included:

questionnaire responses from several female workers; interviews with female workers

and stakeholders in the DCI, public documents and even theory itself. As noted in

Chapter 3, Layder’s Adaptive Theory (1998) suggested that engaging with multiple

sources of data or evidence, adds “strength, density and validity of theoretical ideas and

concepts that emerge from the data collection and analysis” (Layder, 1998, p. 68).

Layder specified three types of data: empirical (emergent or extant) and theoretical.

Theoretical data includes “different forms, types and levels of theory” (Layder, 1998, p.

177), where existing theory scaffolds the research. Multiple sources of data are congruent

with the strategy of enquiry of case study and the underpinning philosophy of Critical

Realism.

In the conduct of the case study, multiple sources of data were collected,

including empirical primary data from participants and empirical secondary data from a

range of sources, including industry documents and popular media. Following Layder’s

(1998) definition (as presented in Table 20) the questionnaires and interviews with the

female interactive content creators form the empirical ‘emergent’ data; as does the semi-

structured interviews with the stakeholders. Layder (1998) suggested that emergent

research data focuses on the immediate findings and information stemming from a

research project. The empirical extant research data includes the secondary data, such as

the government reports and industry websites. As Chapter 3 introduced, extant theory

may also be a source of data (Layder, 1998, p. 72). Aligning to Layder’s approach the

development and use of Framework 2–the Human Agency Multi-Theory Scaffold

(HAMTS)–introduces a body of theory, which can be used as theoretical data.

51 Several women in the multimedia industry could provide certain insights into the games sector, as they knew of someone in the games industry or had considered the career pathway themselves.

Table 20 Types of data in the case study based on Layder’s Adaptive Theory (1998) Layder’s definition As implemented in the case study Empirical 1. Emergent research data -Online questionnaire and semi-structured

interviews of women working in the DCI -Semi-structured stakeholder interviews

2. Extant research data -Secondary data sources; for example, documents (See Appendix 17 for full list)

Theoretical 3. Theoretical data; includes pre-existing general/substantive theory

-Theories in the Multi-Theory Scaffold (as outlined in Chapter 2)

Table 21 summarises the data collection, which included interviews with 18

female interactive content creators and 3 stakeholders. Most of the data collection

occurred over two distinct periods. In Phase 1, of the 13 female interactive content

creators who completed an online questionnaire (See Appendix 8), 12 were subsequently

interviewed. In Phase 2, a further six female interactive content creators were

interviewed. Phase 2 participants did not complete the questionnaire, primarily because

the tool no longer existed. Upon reflection, it may have been useful to continue its use.

Furthermore, 3 stakeholders were interviewed in Phase 2. Thus, the 21 interviews formed

the primary empirical data. In addition, a range of secondary data (See Table 22 for

examples and Appendix 17 for a complete list) offered a way to triangulate the primary

data to add further richness to the research findings.

Table 21 Data source and collection method employed in the reported case study Phase/Year Data source Data collection method

online questionnaire

semi-structured interview

secondary data

Phase 1 (2007)

Female DCI professionals

13 responses * 12

Phase 2 (2009)

Female DCI professionals

Not administered 6

Industry stakeholders

Ongoing throughout

Secondary sources

Total 13 questionnaires 21 interviews (See Table 27 for complete list)

over 60 items (See Appendix 17 for complete list)

* One female interactive content creator’s interview was not recorded due to equipment failure

1) PRIMARY DATA : ONLINE QUESTIONNAIRE

An online questionnaire appeared to be an appropriate method to collect initial

data from the female DCI workers. There were several considerations in the design and

administration of the questionnaire employed in this investigation, which included: 1) the

purpose of the questionnaire, 2) the choice of questions, and 3) the practicalities of

administering the questionnaire.

The questionnaire’s two purposes were: to indicate participants’ commitment,

and to form a tool for the subsequent interview. The primary contact between the

researcher and potential participants was email communication. An initial email provided

participants with a consent form, an overview of the research, and a link to a

questionnaire. Completion of the online questionnaire by a participant indicated their

commitment to be interviewed. Participants completed the online questionnaire prior to a

subsequent face-to-face interview and participants’ responses sensitised the researcher to

each participant’s experiences. Participants’ questionnaire responses were added to the

question bank, which could be used in the subsequent interview (See Appendix 9) if

required; for example, to prompt the participant. Witzel (2000) recommended that

questionnaires be used prior to an interview to sensitise the researcher; however, as Flick

(2009) reminded us, such questionnaires are intended to support the interview process

rather than confine it. Consequently, the research approach remained inductive and

exploratory.

The questionnaire was comprised of both closed and open-ended questions. The

closed questions aimed to capture demographic information such as age, organisation,

and occupational role. This demographic data indicated to the researcher the suitability of

the participants for the study; for example, were they employed in a suitable occupational

role and in their early career stage, and thus fitting the target sample. Furthermore, such

demographic information assisted in the description of the context; for example, the type

of organisation the participant was employed in. The open-ended questions asked

participants about their participation experiences in the workplace. These open-ended

questions were established as a result of the literature review. (See Appendix 8 for the

online questionnaire provided to participants.)

There were several practicalities involved in administering the questionnaire,

which included encouraging participant response rates and the storage of confidential

data. Because a short questionnaire can encourage participant response (Aaker, Day, &

Kumar, 1998) the questionnaire was limited to 7 questions, in addition to questions about

typical demographic data; for example: name and level of education. As ease of access

can also encourage response, participants were emailed a link to the online questionnaire.

This use of email and online technology was seen as suitable because participants

worked in an occupation where technology skills were central. There were several

protocols employed for the storage and confidentiality of data, such as de-identification

of all correspondence (See Appendix 7 for these protocols).

2) PRIMARY DATA : SEMI -STRUCTURED INTERVIEWS

In the conduct of the case study, the primary data collection method was semi-

structured interview of female DCI workers and stakeholders. Interview is a data

collection technique often used in case studies (Yin, 1994, p. 84), in qualitative research

(Creswell, 1998) and in previous research of female participation in IT (Barker &

Aspray, 2006). Furthermore, interview is particularly congruent with research

underpinned by a Critical Realist ontology. Interviews provide insight into informants’

words and views (Miles & Huberman, 1994; Patton, 1990) and, hence, can capture the

individual subject’s insights valued by Critical Realism. As interviews can elicit

participants’ interpretations of their actions and the events they experience (Walsham,

1995b), they can also provide insight into Bhaskar’s (1989) actual and real domains.

Critical Realism requires that explanations be based on participants’ accounts of their

concrete experiences.

In the conduct of this study, phase 1 provided 13 questionnaires and 12

subsequent interviews with female DCI workers. Phase 2 provide a further 6 interviews

with female DCI workers but with no prior questionnaire. These phases and interview

details are summarised in Table 21. As will be discussed, detailed question banks were

developed for the interviews with participants; however, an open approach was

employed with the first question always being ‘Tell me about your experiences of

participating in the DCI’. Although each interview began in the same way, new

questions were often introduced in response to interviewees’ comments or when

interviewees needed to be prompted. These questions included those from the question

bank (See Appendix 9) and those that emerged from the respondents’ questionnaire.

Appendix 9 provides reflective insights on the questions used in the interview.

Stakeholder interviews took a more open-ended approach, with no question banks.

Rather each stakeholder was simply asked for their opinion of women’s under-

representation in the DCI.

In the conduct of this study, semi-structured interviews were employed because

their flexibility is required in exploratory research, where the focus is on discovery in an

emergent rather than tightly prefigured manner. Semi-structured interviews also allowed

for flexibility in the questions asked during interviews. In this way, the researcher could

be responsive to emerging insights and lines of enquiry. Interviews can take different

forms, including structured or open-ended forms (Fontana & Frey, 1994). Semi-

structured interviews (Myers & Newman, 2007) are seen as offering more flexibility than

structured interviews, but not as much as open-ended interviews (Silverman, 2009).

Interviews can also have a particular style. For example, Fontana and Frey (1994)

recommended fostering an interview style where the interaction feels like “real”

conversation (p. 139), and allows a rapport to be established between the researcher and

participants. Accordingly, the researcher employed a personal, friendly interview style.

Specific techniques included those recommended by Holstein and Gubrium (1995), such

as the use of a supportive, empathetic voice which conveys “friendship”. The researcher

found it relatively easy to establish a rapport with participants. This may be because both

the participants and the researcher were female, and members of the DCI community.

The researcher may have, therefore, been considered to be what Patton (1990) described

as a “native”, not an “outside observer” (Trauth, 1997; Walsham, 1995). Furthermore,

the researcher encouraged the participants to ‘tell their story’, as this seemed to be a

personal approach to collecting the ‘data’. Thus, narrative techniques, such as “framing

the interview” and “inviting” the participants’ comments (Chase, 2008, p. 73) were also

used during the interviews.52 This approach resulted in participants’ sharing rich personal

accounts of their participation in the DCI.

Several factors can influence the responses of the interviewee. For example,

participants may “try to ‘assist’ the researcher by giving responses they believe he or she

wants” (Giddens, 1993, p.21). In gender-related studies, the influence of second order

expectations on interview responses has been noted by Giddens (1993). Accordingly, the

research approach employed several techniques to mitigate this influence; these included:

recognising the motivation of the interviewee, using an unbiased questioning style, and

ensuring the clarity of the interview questions. The personal style of interview was

tempered with constant researcher reflection to ensure unbiased data was collected.

The 18 female interactive content creators and 3 stakeholders interviewed

provided approximately 250 A4 pages of data for analysis. The interview protocol (See

Appendix 7) recommended an interview length of approximately 45 minutes; however,

in the conduct of the interviews, the actual length was contingent on the availability of

the interviewee and ranged from 38 to 120 minutes. The digital recording of both audio

and video of all 21 interviews provided an accurate record with which to establish

credibility of data. Video adds credibility and richness to the collected data as it allows

for analysis of non-verbal communication (Spiers, 2004; Gibbs et al., 2002). During the

transcription of the interviews, the researcher could visually establish the emotional state

of the interviewee and accordingly include this in the interview transcript. For example,

one participant, in recounting a difficult personal situation in the workplace reflected

silently for a moment; video captured this moment more accurately than audio could

have done. All interviews were transcribed (in Phase 1 by the researcher and in Phase 2

by a transcription service). Transcripts also captured details such as length of interview 52 There were several texts that guided the researcher on techniques to draw out a participant’s story. These included: James Spradley’s (1979) The Ethnographic Interview,and Gubrium & Holstein's (Eds.) Inside Interviewing.

and quality of audio. They were emailed to participants for verification; this contributed

to the credibility of the research and to what Yin (1994) referred to as a ‘chain of

evidence’ (See Section 4.4.3 for further discussion).

3) SECONDARY DATA

Collection of secondary data followed an approach typical of qualitative research

with data including: documents, direct observation, and physical artefacts, which are also

typical of case study (Stake, 1995; Yin, 1994). As noted in Table 22 , empirical extant

data collected for this case study includes: secondary data sources such as industry

reports, brochures from industry conferences, popular media, raw data from previous

industry surveys, policy documents, promotional materials, and website blogs. One key

piece of secondary data is the international IGDA survey by Gourdin (2005), which

provides over 1000 raw or unedited comments from games workers, including those in

Australia. Appendix 17 provides a detailed list of secondary data sources collected.51F

Media articles, including those appearing in industry journals, form a pertinent

source of evidence. Although it is recognised that introducing media articles as data

sources raises a legitimate concern regarding ethical academic rigour (because such

extant data may not have the validity of empirical data), the use of “popular” literature is

acceptable (Layder, 1998, p. 172). For example, several researchers, including Sumner

and Niederman (2002), refer to the trade publications Computerworld and Information

Week. This type of evidence can, at least, be considered an evidentiary tool (Popora in

Lopez & Potter, 2001, p. 262). As Puth (1996) suggested, “in many cases, existing

secondary data may be sufficiently relevant and comprehensive to answer at least a

certain part of the overarching research question” (p.86).

The primary purpose for including secondary data is to triangulate the subjective

insights of participant interviews. For example, if a participant were to mention maternity

leave (an influence identified in previous research), then a suitable secondary data source

may be maternity leave policies (within the organisation) and legislation (at the societal

level). Secondary data strengthened the researcher’s interpretation of participants’

subjective accounts. Where subjective accounts contrasted with secondary data, these

inconsistencies revealed interesting avenues of further analysis. Furthermore, as Chapter

3 introduced, Critical Realism requires researcher objectivity, and secondary data can

help refine the researcher’s interpretation of the participants’ accounts. In this way,

secondary data is a “resource” (Layder, 1998, p. 165) that can help support emerging

53 The IGDA diversity survey provided over 1000 pieces of raw data from an international survey of games development workers.

lines of enquiry, add richness to participant accounts, and add depth to findings and

emerging theoretical conceptualisations (Layder, 1998, p. 71).

4) THEORETICAL DATA

Layder (1989) suggested that all data, including existing theory, is equally

important for theory development. Thus, the Human Agency theories in the Human

Agency Multi Theory Scaffold (HAMTS)(detailed in Chapter 3) provide a form of data.

However, even though the existing Human Agency theories (as specified in the HAMTS

in Table 11) feature significantly in this investigation, the empirical data (e.g.

questionnaire and interview) remains central for two reasons. First, there is a call for

empirical data in existing research, particularly “for qualitative research that explores

women’s actual experiences” (Adam et al., 2002). Second, collecting subjective insights

(from the individuals involved in the phenomenon under study) is central to the

underpinning Critical Realism ontology because it is only possible to understand social

phenomena by recording and analysing the associated events that take place as a result of

individuals “acting” (Easton, 2009, p. 6). Accordingly, participants’ accounts form the

bulk of the data collected for this case study.

Table 22 Examples of secondary sources of data and reasons for their inclusion (Full list in Appendix 17) Category Evidence Data type Description Purpose for inclusion

Government documents

ABS statistics

Documents (archival)

ABS Survey of Work in Selected Culture and Leisure Activities

Provides objective data such as actual participation rates (2001, 2004, 2007)

Legislative Documents, Workplace Relations Amendment (Transition to Forward with Fairness Act 2008); includes the ten National Employment Standards (NES)

Indicates legislative measures towards parental leave; also highlights changing nature of influences (such as legislation introduced in 2010)

Policy Documents House Standing Committee on Employment and Workplace Relations; inquiry into pay equity and associated issues related to increasing female participation in the workforce (2006)

Suggests ‘influence’ is recognised at a macro level (Participant M2 mentioned male/female pay discrepancy)

Industry research

Survey Documents Game Developer Demographics: An Exploration of Workforce Diversity; reported comments (IGDA, 2005a)

Provides 1000+ unedited comments from the survey; comments are identified for gender, age, disability, and sexual orientation

Interviews Documents Queensland Creative Industry Mapping Study (Higgs & Cunningham, 2008)

Includes qualitative data from 20 face to face interviews of industry professionals employed in Queensland, Australia

Traditional media

Newspaper (print/online)

Documents Sydney Morning Herald blog; comments responding to media article (Appendix 17, Item 16)

Provides insight into how phenomenon is reported to the general public

Promotional material

Documents Flyers promoting the benefits of working at a particular games organisation (Appendix 17, Item 51)

Provide an example of organisational culture; describes working at the company as involving ‘eating pizza and playing games’

Public websites

Industry websites

Documents AIMIA (Australian Interactive Media Industry Association)

Shows typical sites participants can access; sites potential participants might be first directed to

Recruitment Documents Recruitment advertisements (Appendix 17, Item 21) Report a need for employee passion; identify types of occupational roles

Industry events

Documents Websites promoting E3 features, Booth Babes (Appendix 17, Item 7)

Indicate the types of cultural content relevant to ‘gaming and girls’ when using general search terms on the internet

Initiatives Artefact IT Screen Goddess official website and blog (Appendix 17, Item 16)

Exemplifies initiatives responding to phenomenon. Artefact presents women in what may be interpreted as sexualised/glamorous poses.

Book Artefact Tech girls are chic: not just geek (Appendix 17, Item 50)

Exemplifies strategies developed to encourage young girls to consider technology-related careers.

Digital products

Artefacts Computer games Example of male-orientated content in games (e.g. Grand Theft Auto)

Chapter 4. Research Approach ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 105 of 401

4.3 DATA ANALYSIS

As Chapter 3 proposed, employing three different frameworks for the analysis of

empirical data–1) analytical, 2) theoretical, and 3) ontological–can help foster an

understanding of the influences on women’s participation in the DCI.

Framework 1, the analytical framework, emerged from the review of substantive

literature. It fosters analysis of three categories of influences: 1) Person, 2) Environment,

and 3) the Interaction between the two. Framework 2, the theoretical framework, fosters

analysis of the empirical data through the matching of concepts from the agency theories

in the HAMTS(See Section 3.2.4). Framework 3, the ontological framework, stems from

Critical Realism and entails analysis of a stratified reality comprising of three domains of

reality: 1) empirical, 2) actual, and 3) real. Critical Realism also provides the ontology

underpinning the research approach.

Although Critical Realism offers much in the way of an ontological perspective,

it presents challenges in the actual implementation. Danermark et al.’s (2002) Six Stage

Model of Explanation provides a procedural approach that encompasses different forms

of logic in a manner congruent with a Critical Realist philosophy. Several different forms

of logic, or “complementary” modes of inference (Danermark et al., 2002, p. 11) are used

to ensure overall understanding of the phenomenon. Forms of logic include, inductive

and deductive (Dubois & Gadde, 2002), and abductive and retroductive (Danermark et

al., 2002).

In addition, the data analysis utilises established qualitative data analysis

techniques. Although Danermark et al. (2002) provided an approach consistent with

multi-strategic data analysis (pp. 151-154), there is less guidance offered regarding

specific techniques. Consequently, it is proposed that established qualitative data analysis

techniques can support the analysis process. These methods include narrative,

hermeneutic, thematic and pattern analysis. These different “modes” of analysis (Myers,

1997) and specific techniques or methods of analysis assist in investigating different

aspects of empirical data. Indeed, Mingers, (2001) who proposed multi-methodology for

Critical Realist studies, suggested that investigating different ‘levels’ of reality requires

different research methods (Mingers, 2002, p. 303). Thus, the following section

introduces each one of the six stages of Danermark et al.’s model and explains how

different modes and techniques of analysis, such as hermeneutics, can be used in an

integrated and iterative manner within the stages. Section 4.3.1 describes how

Danermark et al.’s (2002) model helps guide the data analysis and theorisation process.

Section 4.3.2 provides examples of the actual data analysis.

4.3.1 DANERMARK ET AL.’S (2002) SIX STAGE MODEL OF EXPLANATION

Danermark et al.’s (2002) Six Stage Model of Explanatory Research provides a

guiding approach for the analysis of empirical data, and one that is highly congruent with

Critical Realism. The six stages are as follows.

1. Description 2. Analytical resolution 3. Abduction/theoretical redescription 4. Retroduction 5. Comparison between different theories and abstractions 6. Concretization and contextualisation

As Table 23 presents, the first 4 stages of Danermark et al.’s (2002) model

loosely align to the three frameworks proposed in Chapter 3. Stages 5 and 6 entail

linking emerging theory to empirical data and align with theorisation reported in the

discussion chapter. However, as Danermark et al. (2002) emphasise, the model is not a

“template to be followed to the letter” (p. 109); the six separate stages can be

“intertwined”, not needing to follow each other in a strict chronological order (p. 109).

Table 23 Danermark et al.’s (2002) six stage model broadly aligned to the three frameworks Stage Framework Framework

Emphasis Logic

1 Description Framework 1: Analytical framework

Identification of E, P, I entities from the empirical data

Inductive 2 Analytical resolution

3 Abduction/ theoretical re-description

Framework 2: Theoretical framework MTS

Use of existing theory to further understand the empirical evidence

Abductive

4 Retroduction Framework 3: Ontological framework

Identification of underlying mechanisms from the empirical evidence

Retroductive

5 Comparison between different theories and abstractions

Synthesis of frameworks

6 Concretization and contextualisation

1) STAGE 1: DESCRIPTION

Stage 1 of Danermark et al.’s (2002) model entails the researcher utilising an

inductive logic to analyse empirically derived data to illuminate participants’

experiences, resulting in a description of the phenomenon. Danermark et al. (2002)

stipulated that an “important part of this description is the interpretations of the persons

involved and their way of describing the current situation” (p. 109). This description of

situation is important for Critical Realism as it stipulates the contextual circumstances of

causal mechanisms (Yeung, 1997, p. 57). As noted earlier, description is also an

important outcome of case study (Yin, 2011), where a descriptive summary of

participants’ accounts can help in the overall analytical process (Eisenhardt, 1989).

Although Danermark et al. (2002) emphasised description of the phenomenon,

they provided less guidance on specific analysis techniques. Consequently, it is proposed

that qualitative research techniques are useful when seeking to establish descriptive

findings. In qualitative research, description may be thin or thick. Thin description

merely reports “facts, independent of intentions or the circumstances that surround an

action” (Denzin, 1989b, p. 33). Stemming from ethnographic approaches (Geertz, 1973),

thick description requires the researcher to identify “the voices, feelings, actions, and

meanings of interacting individuals” (Denzin, 1989b, p. 83) and the “context, actions and

evolution of an act” (Denzin, 1989b, p. 33). Both the techniques of hermeneutics and

narrative analysis are useful in establishing thick descriptions (in contrast to thin) of the

individuals’ accounts. Like narrative analysis, hermeneutics, as a specific mode of

analysis (Prasad, 2002),, can provide “a window into human experience” (Denzin &

Lincoln, 2008) through “interpretation and understanding” (Prasad, 2002).

Qualitative methods such as hermeneutic analysis appear useful for

understanding less readily identifiable influences. Bhaskar’s ‘real’ domain seeks to

understand underlying mechanisms: to examine “what is going on underneath” (Archer,

1995, p. 102). Hermeneutics may be useful as it encourages uncovering this ‘underneath’

or ‘underlying sense’ (Myers, 1994), albeit from an individual’s perspective. Critical

Realism acknowledges that as social phenomena may have an intrinsic meaning when

experienced by humans, that there are hermeneutic and interpretive components in the

research process (Sayer, 2000, p. 17). However, hermeneutics has been criticised for not

being sufficient for a Critical Realist approach (Mills et al., 2009, p. 254) if it simply

offers an explanation of a subject’s meaning, and not the underlying causal tendencies.

Both Bhaskar (1979) and Giddens (1984) recognised that there are unacknowledged

conditions or unintended consequences which exist beyond the agent’s knowledge.

Giddens, who recognises the value of hermeneutic analysis (Byrant & Jary, 1997, p.11)

for understanding subjective insights, noted that a hermeneutic interpretation of a

participant’s meaning is unable to account for the unacknowledged conditions and

consequences of human action (Craib, 1992, p. 25). Therefore, there are limitations

regarding the usefulness of hermeneutic analysis.

There are distinctions to be made as to which form of hermeneutics is of value.

Traditional hermeneutics is often associated with interpreting textual data (Prasad, 2002)

to explore the subjective meaning that an individual may have. Contemporary and

dialectical hermeneutics consider the individual’s actions and their environment.

Contemporary hermeneutics is relevant to the current case study’s focus on organisations

as it considers “organizational practices and institutions, economic and social structures,

culture and cultural artefacts” (Prasad, 2002). Dialectical hermeneutics is congruent with

the case study’s focus on the interaction of the person and environment because it can

consider both the subjective meanings of individual actors and the social structures that

condition meanings (Darke, Shanks, & Broadbent, 1998).

Both narrative and hermeneutic analysis techniques are useful in fostering

descriptive insights of participants’ experiences, as both value a participant’s subjective

insights. However, both face limitations in identifying the less readily visible

mechanisms that Critical Realism values. Critical Realism aims to understand not only

“what happened but why it happened” (Montano & Szmigin, 2004). Thus, there is a

need to employ further analytical techniques.

2) STAGE 2: RESOLUTION

Danermark et al. (2002) suggested that Stage 2 entails the researcher continuing

with their description, while recognising its specific components. The researcher must

“dissolve the composite and the complex by distinguishing the various components,

aspects or dimensions” (Danermark et al., 2002, pp. 109-110). Danermark et al. (2002)

proposed that these component, aspects and dimensions emerge from the empirical data.

Although it is expected that analysis of empirical data will reveal these components,

aspects, and dimensions, it is proposed that Framework 1–the guiding analytical model–

provides a starting point. The categories in Framework 1 serve as what Layder (1998)

called ‘orientating concepts’ where, although coding may occur around the categories (p.

113), these categories do not necessarily influence the data collection (p. 111); nor must

data fit these categories. The broad categories in Framework 1 simply foster initial

synthesis of a large volume of rich and complex qualitative data.

Similarly, qualitative researchers recognise that description may involve

“summarizing, synthesising and extracting essential features or characteristics of a

situation” (Morse & Richards, 2002, p. 150), where the aim is to avoid ‘garden path

analysis’52F54 (Richards as pers.comm, in Bazeley, 2009). Even for qualitative research,

“description is part of an analytic journey”; however, “it alone may not be sufficient”

(Beazley, 2007). Thus, description can involve analytic techniques to identify categories

and/or themes (Mills et al., 2009, p. 927); the distinction between a theme and a category

is that the former runs throughout the data (Bazeley, 2009), and the latter does not.

This recognition of patterns can involve recontextualisation of emerging

empirical data. Recognising patterns can involve specific analytical tasks including:

contrasting, comparing, cataloguing, and classifying data (Miles & Huberman, 1994, p.

37). For example, narrative analysis encourages a “process theory explanation”, which

provides insight into chronologies, time, and connections within the big picture (Miles & 54 Richards describes ‘garden path analysis’ as the reader being taken "along a pleasant pathway that leads nowhere: ‘Here are the roses, there are the jonquils, and aren’t the daffodils lovely today!'"

Huberman, 1994, p.146). Narrative analysis offers a way of organising experiences of

actions and events over time (Chase, 2008, p. 64) by fostering a life story perspective

(Morse & Richards, 2002, p. 118). However, from a Critical Realist approach, there are

limitations to using narrative analysis. Although Critical Realism values individual

insights, there is a need to move towards a synthesis of these insights. Narrative analysis

may not achieve group level synthesis as it has a focus on the “uniqueness of each human

action and event rather than their common properties” (Chase, 2008, p.65). The

challenge lies in understanding both individual accounts and group level analysis of

women’s experiences.

Patterns and relationships also emerge in case study from the aggregation of

codes or instances in data (Stake, 1995, p. 74). Analytical techniques, such as pattern

analysis, can help in the identification of entities at both the individual and group level.

This group level synthesis is integral to Critical Realism, which recognises that whilst

there are “multiple subjective realities” or “multiple realities” (Lincoln & Guba, 1985),

ontologically, there is synthesis towards an objective reality. This offers, as Sayer (2008)

suggested, not so much “multiple realities” as “multiple aspects of a single reality” (p.

11). Thus, it is proposed that analysis follow a dialectic and iterative process that

involves data dissembling, the examination of patterns and relationships, and its

reassembly to explain. Such patterns are identified by examining the data:

1. within each participant’s data set (which includes questionnaire, interview,

and secondary sources of data stemming from participants’ accounts)

2. across participants’ data sets

3. against secondary data

Applying Stages 1 and 2 of Danermark et al.’s (2002) model encourages a

description of the phenomenon under study. Stage 1 fosters a description of the entities

and context (for example, environment and person characteristics involved in

interaction). In Stage 2, description can be further refined and focus on events, as the

process of conceptual abstraction includes “isolating an essential aspect of a concrete

course of events” (p. 43). The approach of using pattern interpretation can assist a

researcher to identify “critical, influential, or decisive” events (Miles & Huberman, 1994,

p.115). Recognition of these events is integral to the ‘actual’ domain that Critical Realists

seek to understand (Bhaskar, 1978, 1989). These events can be interpreted even if, as

Mingers (2004) pointed out, participants themselves may not be aware of these events.

Although description of a phenomenon is a valuable research outcome of

research, Mingers (2002) reminded us that a Critical Realist is never content just with

“description” (p. 302). Description is just the “first step” before hypothesising the

structures and mechanisms that shape observable events (Mingers, 2002, p. 303).

Although the researcher may well become sensitised to possible mechanisms whilst

conducting Stages 1 and 2, there is no explicit analytical focus on the mechanisms.

Accordingly, Stages 3 and 4 move the analytic process beyond description and towards a

specific focus on investigating the underlying mechanisms.

3) STAGE 3-ABDUCTION OR THEORETICAL REDESCRIPTION

Danermark et al. (2002) proposed that Stage 3 requires the framing and reframing

of the empirical phenomenon within differing theoretical frameworks. In Stage 3,

“several different theoretical interpretations can and should be presented, compared and

possibly integrated with one another” (Danermark et al., 2002, p. 110). The objective is

the identification of a framework that best explains the phenomenon and/or results in

new ideas and, through employing the logic of abduction, shows how something might

be (Danermark et al., 2002). Danermark et al.’s (2002) conception of abduction differed

from Peirce’s (1955) conception; the latter emphasised abduction from empirical data

and not from pre-existing theory53F

In this investigation, the use of existing theory during data analysis (using

Framework 2) is not dissimilar to pattern matching, an analytic strategy recommended

for case study by Yin (2004). However, pattern matching utilises a deductive logic (Yin,

2004) to match a predicted theory pattern to empirical data. Danermark et al.’s (2002)

Stage 3 requires that an abductive logic be employed. An abductive logic does not

require empirical data to fit any pre-existing concepts or theoretical pattern but, rather,

should remain open to emergent codes.

A similar argument is made in qualitative research, where Bazeley (2009)

suggested that “there is no problem with a priori categories or themes as long as they are

recognised and declared as such, and they are actually supported in the data; the analyst

can still retain flexibility and be open to the presence of finer nuances or different

emphases in the data” (n.p). In this way, the theories used in the abductive stage do not

constrain the data analysis; they may indeed, “introduce new ideas” (Danermark et al.,

2002, p. 96). Although Danermark et al. (2002) recognised the value of different

theories; they do not offer an explicit method for utilising theories or multiple theories.

Chapter 3 proposed an approach that employs ‘different types’ of theories (such as meta,

critical, middle range and operationalised) to scaffold research, where different types of

theories offer different levels of abstraction.

It is proposed that different theories in the MTS (used to develop Framework 2)

can reveal different mechanisms. In addition, different theories can also describe “the

same mechanisms in a different way” (Danermark et al., 2002) and thus allow the 55 ‘Abduction’ is a term coined by Peirce. Abduction allows us to “understand something in a new way by observing and interpreting this something in a new conceptual framework” (Danermark et al., 2002, p. 77).

researcher to consider a theory’s “validity and explanatory power” (p. 147). Such

multiple perspectives are congruent with a Critical Realist “logic of explanatory

pluralism” (Reed, 1997, p. 32).

4) STAGE 4: RETRODUCTION

Danermark et al. (2002) proposed that Stage 4 transcends deduction, induction

and even abduction (pp. 85-95). Retroduction being an intuitive and creative process,

rather than a logical one (Mingers, 2006). If abduction draws on existing theories for new

insights, retroduction returns to abstracting mechanisms from the data. The analytical

move is from a description of the concrete phenomena towards explanation by asking

what produces them or is a condition for them (Bhaskar, 1986, p. 11, Note 26).

Danermark et al. (2002) described retroduction as advancing “from one thing (empirical

observation of events) and arriving at something different (a conceptualization of

transfactual conditions)” (p. 96). Retroduction is an important aspect of a study that

adopts a Critical Realist perspective, as it is “a key form of thought operation in

theorising and theory generation” (Danermark et al., 2002, p. 149).

Accordingly, Stage 4 may involve further identification or ‘abstraction’ of the

underpinning mechanisms involved in the interaction between the individual and their

environment. Mingers (2006) described this as ‘hypothesising’ the mechanisms.

Although mechanisms may not be readily observable, for the Critical Realist, the causal

effect implies existence, regardless of observability. Thus, retroduction, which involves

counterfactual thinking (Danermark et al., 2002, pp. 80,101), may lead the researcher to

ask questions such as: How would this be if not...? Could one imagine C without ...?

(Danermark et al., 2002, p. 101). In regards to specific analytic techniques, pattern

analysis is useful in synthesising mechanisms (occurring in the real domain) that

underpin the experiences (in the empirical domain) and events (in the actual domain).

However, as will be discussed in the findings and discussion (Chapters 5 and 6,

respectively), the abstraction of mechanisms provides a challenge for the researcher.

5) STAGE 5: THEORIES AND ABSTRACTION

Danermark et al. (2002) proposed that Stage 5 elaborates and estimates the

relative explanatory power of the mechanisms and structures identified in the previous

stages (p. 110). The emphasis during Stage 3 (abductive) is on the use of theories to

illuminate possible underlying mechanisms and, in Stage 4, (retroduction) mechanisms

are identified by the researcher interpreting the empirical data. In Stage 5, the analytical

process moves further towards explanation by synthesising the previous stages (for

example, theory and empirical evidence). This stage also involves a refinement of which

specific mechanisms the study considers because, as Sayer (1992) reminded us, “even

when we are interested in wholes we must select and abstract their constituents” (p. 86).

Synthesis also further moves the case study towards the specification of an emerging

theory. As Section 4.5 further argues, there are considerations to be made in regards to

the type of theorisation that emerges from the research findings.

6) STAGE 6: CONCRETIZATION AND CONTEXTUALISATION

Danermark et al. (2002) proposed that Stage 6 involves examining “how different

structures and mechanisms manifest themselves in concrete situations” (p. 110). Hence,

further analysis takes the ‘mechanisms’ abstracted from the data and theory (in the

previous stages) and ties them back to the empirical data. As will be discussed in the

findings and discussion chapters (Chapters 5 and 6, respectively), the emphasis is on

illuminating how the mechanisms identified during analysis interact and manifest in a

concrete situation.

4.3.2 ANALYSIS OF DATA

As noted in the previous section, Danermark et al.’s (2002) model guides the data

collection and analysis. Reflecting stage 1 of Danermark et al’s model, the research

process takes an initially open approach to collecting data that would illuminate people’s

actual experiences. Even though tools such as questionnaire’s and question banks were

developed the approach remained inductive. By stage 3 there is a different logic at play

that of abduction, and the data analysis draws more on theory (such as the HAMTS). By

stage 5 and 6 there is perhaps less coding of data as the focus moves to the development

of theoretical linkages, including the identification of elements and relationships.

The software package NVivo was utilised most strongly in the early stages of

coding, in most to manage the volume of rich data. Following data collection, and prior

to the development of the three framework, the coding followed a process of refining a

large number of free nodes to a smaller set of inductively derived tree nodes. In this early

stage it became evident that the data analysis could benefit from an analytical framework

and ultimately the literature led to the development of the categories ‘E, P and I’ in

Framework 1. The large volume of data was then re-analysed using these categories.

These three initial categories or nodes were further refined to be the analytical categories

presented in Figure 9 and eventually the Sphere of Influence model (Figure 23) - the

model resulting from the analysis of the data using Framework 1.

Figure 8 Example of Nvivo project showing tree nodes

During the coding process it became obvious that not all data would ‘fit’ the

‘E,P, I’ model (being Framework 1) and this lead the researcher to consider other

analytical categories and to eventually propose further frameworks. For example,

participants described the events that were memorable to them, the events that influenced

their participation. The researcher created a node titled ‘events’, which sat outside of the

early nodes. At a much later stage, when the third framework was introduced the ‘events’

node was moved to sit under that framework. Framework 3 is the critical realist

approach, which seeks to identify ‘events’. Similarly, participants often described their

confidence and the researcher reflected on the value of introducing self-efficacy theory to

further understand the concept and ultimately Framework 2- the Human Agency Multi

Theory Scaffold was developed. Data was the recoded to the categories that framework 2

provided. This is further discussed in the findings chapter in Section 5.3.

In the conduct of this research, the process of analysis involved employing

several methods such as thematic mapping, pattern matching and hermeneutic analysis,

in a highly iterative and integrated manner to code the data across the three frameworks

proposed in Chapter 3. Table 24 provides a summary of the different coding activities,

data sources (Section 4.2.2), and techniques used in data analysis (Section 4.3) as

relevant to these three different frameworks (presented in Chapter 3). Although Table 24

aligns the three frameworks to Danermark et al.’s (2002) model, the actual process is

highly iterative and does not necessarily have such distinct analytical boundaries.

Table 24 Summary of key analysis activities aligned to the three frameworks used, and to Danermark et al.’s (2002) model of explanation Framework Coding activities Data source Domains of

reality (Bhaskar, 1978, 1989)

Stage of Danermark et al.’s (2002) model

Example of data analysis techniques

Example outcome of analysis 56

Identify key categories in framework

Categories emerge from initial review of literature

2. Resolution Identify three key categories of environment, person and interaction

Identify key themes emerging within and between accounts

Interviews, secondary data

Empirical (individuals’ experiences)

1 Description

Thematic, pattern

Identify lifespan perspective with which to contextualise interactions

Code for descriptions of the individual

Empirical 1 Description Hermeneutic, Narrative, pattern

Identify tensions between gender and occupational identity

Code for descriptions of the environment

Empirical 1 Description Hermeneutic, narrative, pattern

Conceptual framework of the environment identifies temporal aspects and macro and micro contexts; for example, organisation

Code for events of interaction

All empirical data recoded to the E, P, I categories

Actual (events and actions)

2. Resolution Hermeneutic, narrative, pattern

Events such as ‘motherhood’

al Code for the

categories identified in the Human Agency MTS, e.g. self-efficacy

All empirical data recoded using concepts in HAMTS

3. Abduction

Pattern Draws on agency theory for analysis and explanation; helps identify mechanisms, for example, Bandura’s mechanism of self-efficacy

Abstract mechanisms involved in interaction

All empirical data recoded (narrowing focus to mechanisms)

Real (underlying mechanisms)

4. Retroduction Pattern Abstraction leads to heuristic models identifying agent-driven mechanisms

Synthesis (involves reanalysis and reassembly of data)

All empirical data recontextualised to emerging theoretical findings

5 and 6. Theories and abstraction, concretization

Emerging conceptual/theoretical concepts illustrated by empirical examples

56 (written post analysis)

4.4 QUALITY

The quality of the research findings can be, to some extent, established by

communicating the methods. Accordingly, this dissertation clearly articulates all aspects

of the research process. In addition, specific criteria are considered for qualitative

research in general (See Section 4.4.2 ) and for case study in particular (See Section

4.4.3), and the research is underpinned by a Critical Realist approach (See Section 4.4.1).

Addressing quality aspects helps establish a systematic and defensible research approach

and strengthens the resultant findings.

4.4.1 CRITICAL REALIST CRITERIA

Even though no specific criteria for quality has been developed for Critical

Realist research (Healy & Perry, 2000), four measures of quality are considered central:

1) objectivity, 2) triangulation, 3) explanation, and 4) researcher reflectivity.

1) OBJECTIVITY

Sayer (2000) believed that objectivity is imperative for Critical Realist research,

where this objectivity includes: 1) epistemological objectivity, 2) objectivity of the

researcher, and 3) recognition of the value-ladenness of objectivity (p.58).

Epistemological objectivity entails the researcher acknowledging given, real worlds and

the subjects’ beliefs about them, and in so doing, not committing the ‘epistemic fallacy’

(Bhaskar, 1978; Sayer, 2000; Archer, 1995) of conflating ontology and epistemology.

Conflation occurs when ‘what we think is’ (epistemology) is confused with ‘what is’

(ontology). Archer (1995) recommended that ‘analytical dualism’ assist the researcher in

analytically separating elements54F

57. Although analytical dualism involves a form of

bracketing, it differs to Giddens' ‘methodological bracketing’ (Danermark et al., 2002, p.

169). Researcher objectivity entails the researcher reflexively recognising the value

ladenness of research.

Several techniques were employed in the conduct of the case study to achieve

researcher objectivity, such as memoing. Memoing the research process offered the

researcher an opportunity to reflect on the findings and to ascertain if these findings were

biased in any way by personal beliefs. A further form of objectivity was achieved by

initially investigating entities such as the environment and person in a bracketed manner;

this aligned with the analytical dualism required by Critical Realism.

57 Archer’s (1995) own work separates structure (in particular, culture) and agency.

2) TRIANGULATION

As Yeung (1997) noted, Critical Realism places an emphasis on triangulation to

strengthen the quality of findings. Ontologically, Critical Realism proposes that there is a

reality we may know; however, epistemologically, our knowledge claims of it are fallible

(Lopez & Potter, 2001), but not equally so (Yeung, 1997; Danermark et al., 2002).

Therefore, claims about reality must be subjected to the widest possible critical

examination to facilitate comprehending reality as fully as possible.

Triangulation can occur with investigators, data, theories, and methodologies

(Denzin, 1989; Snow & Anderson, 1991; Stake, 2000; Patton, 1990). Data triangulation

is a recommended technique for qualitative research (Denzin, 1989a), case study (Stake,

1995) and Critical Realism (Yeung, 1997). Eisenhardt (1989) commented that, in case

study, this type of tactic “exploits the unique insights possible from different types of

data collection” (p. 541). Triangulation can develop “in-depth understanding, not

validity” (Denzin, 1989a, p.246). Theoretical triangulation asks the researcher to “be

aware of the multiple ways in which the phenomena may be interpreted” rather than be

“consistent” with theories (Denzin, 1989a, p. 246).

In the conduct of the case study, both data and theory triangulation were

employed. As discussed earlier (in Section 4.2.2), data triangulation involved the use of

different or multiple sources of data/information, including participant accounts

(questionnaire and interview), stakeholder (interviews) and secondary data sources

(documents/artefacts). Theory triangulation involves using the four different agency

theories in the HAMTSfor the interpretation of the phenomenon (detailed in Section

3.2.2). Layder (1998) suggested that a multi-strategy approach, which incorporates both

theory and empirical data, “automatically contributes to triangulation” (p. 68).

3) REFLECTIVITY

Critical Realism does not deal with measurable or even observable cause and

effect relations, but rather with interpreting underlying causal tendencies (Bhaskar 1978;

Tsoukas 1989). Thus, data interpretation may be particularly open to researcher bias or to

“the subjective nature of knowledge” (Dobson, 1999, p. 268). Similarly, qualitative

researchers recognise that “value-free interpretive research is impossible” (Denzin,

1989a, p. 23); thus, there is a need for “explicitness about inevitable biases that exist”

(Miles & Huberman, 1994). Critical Realism requires awareness and transcendence of

biased beliefs and a “continual commitment to caution, scepticism, and reflexivity”

(Dobson et al., 2007), where these increased levels of researcher reflexivity can minimise

bias. Similarly, case study encourages researcher reflexivity (Mills et al., 2009).

Accordingly, this investigation employed techniques to minimise researcher bias.

These techniques included adopting tenets of hermeneutic analysis (See Appendix 10 for

those recommended by Klein and Myers, 1999) which require the researcher to be

“deeply self-reflexive and self-critical” (Prasad, 2002). Further techniques involved

articulating the researcher’s position (Myer, 1994); this was done by introducing the

researcher’s background in Chapter 1. In addition, techniques such as Van Manen’s

(1990) “bias bracketing” were useful; these involved the researcher acknowledging and

putting aside her personal beliefs, particularly when confronted with statements during

interviews that challenged these beliefs. Bracketing involves “suspending one's various

beliefs” (van Manen, 1990, p. 175) and making explicit our prior ideas, biases, or

preferences and constant reflection during the research process regarding how to

minimise these possible biases. This differs to Archer’s analytical dualism or Giddens’

methodological bracketing, which are of a more ontological nature. At the same time,

personal involvement was not totally negated, particularly as it serves to fuel the

researcher’s commitment (as noted by Schon, 1983). 55F

In addition, the process of analysis drew on several methods to represent data

because, as Miles and Huberman (1994) suggested, representations can help “make

sense” of data (p. 91). As Appendix 14 indicates, data representation in this investigation

included the use of matrices to organise the data. A high level of reflectivity can be both

maintained and evident in self-reflective accounts in tools such as a researcher’s diary

(See Appendix 13: Example of researcher reflection).

4) EXPLANATION

When using a case study approach, an explanation of a phenomenon may involve

analytic generalisation (Yin, 1984; 1994) as a “penultimate criterion” of theory building

(Healy & Perry, 2000). However, the limitations of such generalisability have been noted

for case study ( Tsoukas, 1989; Flyvbjerg, 2006; Easton, 2003). Indeed, there has been a

“general scepticism about the possibility of explanation” from case study, particularly

when the case study focuses on description rather than explanation (Craib, 1992, p. 26).

Critical Realism holds similar concerns regarding generalisability. While analytical

generalisation from case study is compatible with the Critical Realist focus on causality

(Easton, 2000), this “generalizing may, however, mean different things” (Danermark et

al., 2002, p. 73). Some Critical Realists, such as Sayer (1992), argued that explanations

from generalisation are only acceptable in the absence of knowledge of causal powers.

Others argue that any explanation involving generalising claims must recognise causal

mechanisms (Danermark et al., 2002, p. 1). 58 Schon refers to this as the ‘artistry of practice’

In this investigation, generalisability is not a research goal because, as noted, it is

not congruent with a Critical Realist ontology. However, generalisability can improve the

overall contribution of the case study findings. Accordingly, the findings from the case

study can achieve several forms of generalisation (as summarised in Table 25). This

includes providing rich descriptive insights, linkages between the empirical data and

theory (for example, by using Framework 2), and linkages between empirical data and

existing research (for example, Framework 1). Generalisability can also be achieved with

the resulting findings being applied to other research contexts (See the discussion in

Section 6.2.4).

Table 25 Generalisability of case study findings Author Types of case study generalisability Achieved in the current case

study Walsham (1995a)

Four types: development of concepts, theoretical frameworks, specific implications about the topic of interest and generation of rich insights 56F

Concepts are presented (for example, E and P in the guiding analytical framework); rich description is an outcome

Lee & Baskerville (2003)

Four types: empirical to empirical; empirical to theoretical, theoretical to empirical, and theoretical to theoretical

A theoretical scaffold facilitates generalisations between empirical and theoretical data

Stake (1995); Creswell (1998)

Naturalistic generalization rather than the generalisation of the findings to other settings (Miles & Huberman, 1994).

May be achieved by research providing a narrative account, a story, a logical presentation

Yin (1994; 2004)

Theoretical propositions, where to 'explain' a phenomenon is “to stipulate a presumed set of causal links about it”(Yin, 2004)

No fixed propositions are used; however, theory can be used to explain aspects of data

Tsoukas (1989)

Analytical rather than statistical generalisability57F

60 The use of theory aids analytical generalisability

Maxwell (1992)

Theory extendable to other situations Resultant conceptual frameworks used in other research contexts

4.4.2 QUALITATIVE CRITERIA

TRUSTWORTHINESS

As the case study in this investigation utilises research techniques commonly

associated with qualitative studies, such as thematic and pattern analysis, certain aspects

of Guba’s Model of Trustworthiness (1981) can be used for judging the quality or

“trustworthiness” (Lincoln & Guba, 1985) of findings. Guba and Lincoln (1989)

proposed four criteria: 1) credibility, which corresponds roughly with the positivist

concept of internal validity; 2) dependability, which relates more to reliability; 3)

transferability, which is a form of external validity; and 4) confirmability, which is

59 The dissertation discusses elsewhere that the critical realists underpinnings may involve interpretive approaches. 60 Yin (1993) pointed out that even Giddens failed to distinguish between analytical and statistical generalizability in suggesting that case methodology may be “microscopic” because it “lacked a sufficient number” of cases and, consequently, judgments of “typicality” (Giddens, 1984, p. 328).

largely an issue of presentation (Lincoln & Guba, 1985). Table 58 (See Appendix 6)

summarises the manner in which the research design endeavours to maintain the four

quality criteria and specific validation techniques employed in this study. In

acknowledgment of the pluralistic paradigmatic nature of Critical Realism, the equivalent

measures for positivist research are included in brackets in the left hand column.

4.4.3 CASE STUDY CRITERIA

In addition to the Critical Realist quality concerns (Section 4.4.1) and the

‘trustworthiness’ of qualitative research (Section 4.4.2), there are certain quality

considerations for case study itself.

CHAIN OF EVIDENCE

Maintaining a suitable data management approach is an important aspect of

achieving overall quality in case study research (Yin, 1994). Data management

illuminates how the researcher has progressed from data to findings, thus establishing a

“chain of evidence” (Yin, 1994).

In this investigation, a case study database (in the form of a Nvivo 9 project)

contributes to the overall quality by providing a way to maintain the audit trail. The

database includes the online questionnaire responses, original video footage, verbatim

and annotated interview transcripts, and researcher reflexivity in the form of memoing

and field notes. The database also includes the actual coding categories which were used

for data reduction, as recommended by Conrad and Serlin (2006, p. 418).

PROTOCOLS

A case study “protocol” Yin (1994, p. 64) also contributes to the overall ‘chain of

evidence’ (Yin, 1994; Stake, 1995) to improve the case study’s reliability. In addition, an

interview protocol ensures that the main analytical concerns are addressed in the

interviews (Patton, 1990; Creswell, 1998). In this investigation, the case study protocol

includes: a schedule; an overview of the project; field procedures (for example,

credentials and access to sites); guidelines for the management of data and

confidentiality; and digital encoding requirements (See Appendix 7). The interview

protocols for the current study contain: the main questions, probe questions, location

requirements, length of interview, guidance on self-disclosure, and guidelines on the

style of interview (See Appendix 8 for the interview protocol used for key informants

and stakeholders).

4.4.4 ETHICS

Maintaining quality requires that the study address ethical considerations. For

this investigation, ethical considerations included maintaining participant anonymity and

recognising the researcher’s axiological orientation; these were in addition to the ethical

requirements of the supervising institution. Ethical clearance involved maintaining

participant anonymity, particularly as informants may share sensitive personal

information. However, anonymity itself is a question of degree, being satisfied either in a

weak or strong form (Yow, 1994; Wengraf, 2001). In the weak form, the informant

would not be recognisable by the general public (although they may be recognisable by

people who know them well). A stronger version is one in which the informant would

not recognize himself or herself in the published account (Wengraf, 2001, p.187).

As has been identified earlier, the empirical data collected for the current case

study includes people’s accounts of experiences. To ensure the anonymity of

participants, a weak form of anonymity is adopted. This weak form is achieved by

sufficiently changing certain identifying details (name, place, age, occupation, etc.) on all

data records (video of interview, questionnaire). All participants and organisations are

given unique unidentifiable IDs, which can be used instead of real names in the reporting

of findings. Participant IDs begin with a letter indicating the micro context; for example,

multimedia (M), or games (G). Each letter is followed by a number to indicate the

interview sequence; for example, ‘M1’ is the first participant interviewed from the

multimedia industry. Similarly, organisations in which participants are employed are

coded with a letter denoting the ID (as indicated in Table 23), and a stakeholder is coded

as ‘S’.

Ethical considerations also involve recognising the researcher’s axiological

concerns, as these may determine aspects of the research, such as choice of problem,

choice of the paradigm to guide the problem, choice of theoretical framework” (Denzin

& Lincoln, 2008, p.169). In this investigation, axiological concerns are reflected in the

inclusion of a critical theory in the Multi-Theory Scaffold. Such critical theories sensitise

the researcher to issues such as “social justice, equity, nonviolence, peace and universal

human rights” (Denzin & Lincoln, 2008, p.13).

Even though the current study is concerned with women’s participation, it is not

defined as ‘feminist research’. This follows Calás and Smircich’s (2009) suggestion that

the difference between “feminist” and “non-feminist research” lies in the former

explicitly focusing on the emancipation and transformation of women, and having gender

as the “central axis” (p. 249). However, agency theories themselves do have an

emancipatory aspect. Reed (1997) noted that drawing on agency theories from a Critical

Realist approach introduces emancipatory aspects by providing “well-informed

critiques” of an agent’s “restrictive influence and emancipatory potential” (p. 34). Thus,

even though Giddens is wary of using the term ‘emancipatory’ (Bryant & Jary, 1997),

and Critical Realism has been critiqued for a lack of focus on emancipatory concerns

(Denzin & Lincoln, 2008) or for not being ‘critical’ enough (Hammersley, 2009), it does

have emancipatory concerns (Easton, 2009). Such ethical considerations or ‘tensions’

(Walsham, 2006) must be considered as they can influence research outcomes.

There are also the more personal ethical decisions that a researcher faces when

interacting with participants. As Denzin (1989c) reminded us “… we must remember that

our primary obligation is always to the people we study, not to our project or to a larger

discipline” (p.83). There are further ethical concerns regarding the consequences of the

study’s findings. The reporting of 'gender' issues is itself a part of gender construction

and, hence, research can involve a moral intervention (Giddens, 1976, pp. 8,162). It is

important to recognise that through “slippage” (Giddens, 1976, p. 162), research findings

may enter the everyday vernacular of people. Thus, the researcher must remain aware of

the nature of the material published or distributed by the academic community and the

media.

4.5 THEORY GENERATION

Much of the discussion regarding theory to this point of the dissertation has

focused on the use of existing theory to scaffold the study, in the form of the Human

Agency Multi-Theory Scaffold. There is also a need to consider the development of

empirically-grounded theory, particularly as both case study and Critical Realism

encourage theory development.

A researcher must be aware of the contentious points surrounding theory

development. There are differing “views on what constitutes theory” (Gregor, 2002, p.

2), where even the term ‘theory’ is ambiguous (Hammersley, 1995). Thus, there is a

need to give an explicit definition of the researcher’s view of theory (Gregor, 2002), and

to explicate the methods by which any resultant theory is derived (May et al., 2009).

Accordingly, the following three questions regarding the development of theory are

considered.

1. What form might an emerging theory take?

2. What type of theoretical explanation may result?

3. How might good theory be developed?

1) The first question asks: What form might an emerging theory take? There are different

views on what “theory is or is not” (Sutton & Staw, 1995, p. 371). Theory may comprise

of written statement of entities, constructs, relationships or propositions (Sutton & Staw,

1995). Giddens (1988) distinguished between a theory and a scheme, sensitising

metaphors, and analytic devices (such as analytical models). Such analytical devices can

be used to construct a descriptive scenario of how, why and how events in an empirical

situation transpired” (Giddens, 1988, p. 162). As sensitizing analytical schemes, these

models are a good “starting-point” for building testable theory (Giddens & Turner, 1988,

p. 165).

In contrast to strong theory, tools such as sensitising schemes and typologies aid

theory development and represent the interim struggle (Weick, 1989). Typologies, which

“fall partway between description and theory” (Bazeley, 2007, p. 192), help to order data

and findings (Lofland & Lofland, 1995). Developing typologies requires the researcher

to specify concepts (Whetten, 1989), and help “to generate and stimulate theoretical

thinking by encouraging the researcher to make comparisons between phenomena”

(Layder, 1998, p. 162) and engage in theoretical elaboration (p. 73). Thus, theory

development can involve the development of models and frameworks.

2) The second question asks: What type of theoretical explanation can result from

the current study? As Section 4.1 proposed, both case study and a Critical Realist

approach can achieve description and explanation as an outcome. However, when

answering the ‘W questions’–What, Who, When, Why (Whetten, 1989, pp. 490-492)–a

Critical Realist perspective may emphasise the ‘how and why’ (Jeppson, 2005; Layder,

1998, p. 100). Even though the main essence of theory is to provide an explanation of

observed phenomena, different types of theory may offer different types of explanation

(Gregor, 2006; Llewellyn, 2003). For example, in considering Gregor’s (2006)

taxonomy of five types of goals for theory development, the current study could achieve

theory that is descriptive (Type 1). Type 1 theory offers a form of explanation by

identifying ‘what is’, by describing and categorising entities (p. 6). A stronger

explanation may be offered by what Gregor (2006) calls a ‘Type 2 theory’, which aims to

explain ‘how and why’. Thus, this dissertation can offer different types of theoretical

explanation as an outcome.

3) A third question asks: How might ‘good’ theory be developed? Answering

this question entails recognising that there are important links between methodology,

type of theories generated, and aspects of the empirical experience that are emphasised

(Shoib et al., 2006) and, in particular, between the ontology and the practical theory that

can emerge (Archer, 1995). As a methodological approach, case study provides analytic

strategies for theory building. Although case study can employ useful strategies for

theory development, such as “safe guards” (Yin, 1994, pp. 120-121) and a “road map”

(Eisenhardt, 1989), there are, as Eisenhardt (1989) suggested, “no generally accepted set

of guidelines for the assessment” of theory building from case study (pp. 547-548).

There are, however, criteria that any emerging theory can address, including

providing an organized body of concepts and principles to explain a particular

phenomenon (Leedy & Ormrod, 2005), and in a parsimonious manner (Eisenhardt,

1989). From a Critical Realist perspective, theory should specify concepts and the

relationships between them, while remaining “close to observable data” (Danermark et

al., 2002, p. 126). That is, theory development moves from the concrete to the abstract

(Layder, 1998, pp. 100-101), and then returns to the concrete (Danermark et al., 2002).

This investigation aims for empirically informed theoretical understanding of the

phenomenon under study.

4.6 CHAPTER SUMMARY

Chapter 2 introduced several methodological considerations emerging from the

review of literature relevant to the research problem. Chapter 3 introduced three specific

frameworks with which to undertake data analysis. This chapter has further articulated

the research approach, as summarised in Table 26.

Section 4.1 illuminated five specific reasons why case study, as a strategy of

enquiry, is highly congruent with the research problem at hand. Section 4.3 highlighted

that Danermark et al.’s (2002) Six Stage Model of Explanatory Research is a useful

guiding approach for data analysis along with typical qualitative techniques. The Multi-

Theory Scaffold (MTS), first proposed in Chapter 3, is particularly relevant in Stage 3 of

Danermark et al.’s (2002) model, which involves drawing on theory using an abductive

logic. Section 4.3 also explained how qualitative data analysis techniques are of value

when using Danermark et al.’s. model. Section 4.4 outlined several considerations to

ensure the quality of the research outcomes. These considerations emphasise that any

interpretation requires of the researcher a critical reflexivity to ensure their

presuppositions do not affect the gathering of the data and interpretation of findings.

In contrast to the previous chapter’s emphasis on articulating the use of existing

theory to support the research process, Section 4.5 presented the considerations for

theory development. From a Critical Realist approach, the analytic process of theory

development must remain grounded in the empirical data–beginning and ending in the

“concrete” data (Sayer, 2000). Thus, any emerging ‘explanation’ presented in the

following chapters involves the synthesis of empirical data, theoretical insights and

ontology, where the ultimate aim lies in the abstraction of the underlying mechanisms or

‘causes’ of participation. Having outlined the methodological plan, the following chapter

reports the findings emerging from the case study.

Table 26 Summary of research design presented in Chapters 3 and 4 Research Design

As applied in the current study

Research questions Due to exploratory nature, start broadly and refine through emergent data,

theoretical and philosophical perspectives

Philosophical perspective (paradigm)

Ontology Critical Realism; for example, Bhaskar’s (1986) three domains of reality (real, actual, and empirical)

Theoretical perspective

Substantive theory

Draws primarily on a range of relevant literature–including gender, ICT and sociological literature–to build on cumulative knowledge in an interdisciplinary manner

Methodological approach Strategy of enquiry

Case study (exploratory, descriptive, and potentially explanatory (outcomes)

Object of enquiry

- Individual (P-person) - Environment (E-environment) - Interaction (I) between E and P The primary object of enquiry, or unit of analysis, is interaction

Data collection Multiple sources of qualitative data including: questionnaire, semi-structured interview, documents and artefacts

Data analysis Employs three frameworks for analysis of empirical data (as discussed in chapter 3). 1. analytical- E, P, I (see section 3.1) 2. theoretical- Human agency theory (HAMTS)(see section 3.2) 3. ontological- critical realism (see section 3.3) Data collection, analysis and theorisation follows Danermark et al.’s (2002) Six Stage Model of Explanatory Research (as discussed in Section 4.3.1)

The actual techniques used were in most qualitative (as discussed in Chapter 4) and included a range of established techniques, including hermeneutic, narrative, thematic, pattern analysis.

The types of data included empirical and theoretical (as discussed in Section 4.2.2, where theory is used as a scaffold. The MTS (Multi-Theory Scaffold) comprised of Human Agency theories (HAMTS) is used as a guide and extant data.

Research quality Various strategies, including data and theory triangulation; reflective memoing to minimise researcher bias (see section 4.4)

Theory development

- Draws on Adaptive Theory (Layder, 1998) to utilise existing theory and empirical data for theory development (as discussed in Section 4.2.2) - Uses analytical schemes or conceptual tools to foster theory development - Moves towards explanation, where explanation from a Critical Realist approach requires the identification of causal mechanisms

Chapter 5. Findings ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 125 of 401

Chapter 5: Findings

This chapter presents the research findings that help address the research

problem: ‘Why do women participate in the Australian Digital Content Industry (DCI)?’.

Two research questions were posed in Chapter 2 in reference to the research problem.

The first asked: What are the influences on women’s participation in the DCI? The

second asked: How can we understand these influences?

Chapter 3 proposed that three different frameworks could be used for the analysis

of empirical data: 1) analytical, 2) theoretical, and 3) ontological. Each framework offers

a way to identify and understand the influences on women’s participation, primarily by

considering the interaction between the environment and the person.

This chapter presents the research findings, and the empirical evidence that

addresses the two research questions. Accordingly, this chapter is structured as a report

on the application of the three frameworks used to analyse the empirical data. Although

the research design has been clearly articulated the overall research process has been

highly iterative.

The findings from the application of Framework 1 provide rich descriptive

insights of the environmental and personal characteristics and their interactions, which

can influence women’s participation (See Section 5.2). The influences are categorised in

a manner that reflects the analytical model offered by Framework 1: 1) Environment, 2)

Person, and 3) Interaction between the two. As noted in Chapter 3 (Section 3.1), the

analytical model emerged from the initial review of literature in Chapter 2.

The findings from the application of Framework 2 offer theoretical insights by

drawing on the Human Agency Multi-Theory Scaffold (HAMTS) (See Section 5.3). Using

the concepts from the HAMTS draws to attention influences such as role models, social

norms and self-efficacy. As noted in Chapter 3 (Section 3.2), the theoretical model

emerged from the development of a Multi-Theory Scaffold (MTS) comprised of four

Human Agency theories.

The findings from the application of Framework 3 offer explanatory insights by

identifying underlying mechanisms involved in the interaction of the environment and

person (See Section 5.4). As noted in Chapter 3 (Section 3.3), the third framework is

based on the ontological stance of Critical Realism.

Each framework reveals different types of influences and provides different

approaches to understanding those influences. Many of these influences have been

identified in previous research regarding women’s participation in the ICT industry or

the DCI. The findings provide rich descriptive insights of the influences; illuminating a

deeper of understanding of how and why an influence is at play. The three frameworks

reveal a comprehensive and detailed insight into the influences relevant in the DCI

context.

The key finding emerges from the synthesis of the insights from the three

individual frameworks; however, in-depth discussion of this synthesis is reserved for

Chapter 6. As Chapter 6 discusses, the three different frameworks offer a complementary

insight into a range of influences. The synthesis of the findings fosters the

conceptualisation of agent-driven mechanisms. Agent-driven mechanisms offer an

original approach to understanding women’s participation in the DCI.

5.1 AN OVERVIEW OF THE CASE STUDY

The following brief case study overview begins the process of identifying the

influences on women’s participation by describing the context and participants. A

description of the context reveals that women are under-represented in both games and

multimedia organisations. A description of the participants interviewed reveals

demographic information, such as the fact that only two of the 18 women interviewed

had children. This description reflects Stage 1 of Danermark et al.’s (2002) six-stage

model as it entails a description of the contexts and entities involved in the phenomenon.

An important part of description is also the “interpretations of the persons

involved and their way of describing the current situation” (Danermark et al., 2002, p.

109). Accordingly, the following findings provide rich verbatim quotes from the

interview transcripts. Several quotes are lengthy in order to capture the richness of

participants’ accounts and to add to the credibility of the researcher’s interpretation.

CASE STUDY CONTEXT

The participants interviewed were all employed, at some stage of their careers, in

the Australian DCI and based geographically within the city of Brisbane, Australia. The

organisations they were employed in ranged from small family businesses, to large multi-

nationals, to government agencies. At the beginning of the interview period (2007),

Brisbane was one of the largest hubs for Australian based games development. However,

by the end of the interview period (2011), three games organisations had retrenched a

majority of staff or closed their Brisbane studios. This may have been a consequence of

the two global financial events that occurred over the study period.61

61 For example, the Global Financial Crisis (GFC) in 2008.

DESCRIPTION OF THE PARTICIPANTS

Chapter 4 described the data collection methods, which included an online

questionnaire and semi-structured interviews. The sampling strategy described in

Chapter 4 resulted in interviews with 18 female interactive content creators (See Table

27) and 3 stakeholders (See Table 29).

Table 27 Case study participants (female DCI workers and industry stakeholders) Participant No.

Participant ID *

Position Title Gender Number of pages **

1 M1 Creative Director Female 9 2 M2 Multimedia Developer Female 16 3 M3 Software Engineer Female 9 4 M4 New Media Producer Female 10.5 5 M5 Web Developer Female 12 6 M6 Web Developer/Designer Female 7.5 7 M7 Multimedia Developer Female 13 8 M8 Online News Editor Female 12 9 M9 Interactive Product Developer Female 17 10 M10 Social Media Coordinator Female *** 11 G1 Production Assistant Female 13 12 G2 Junior Props Artist Female 10 13 G3 Artificial Intelligence Programmer Female 13 14 G4 Project Coordinator Female 14.5 15 G5 Assistant Producer Female 22 16 G6 Games Designer Female 13.5 17 G7 Senior Character Artist Female 22 18 G8 Web Designer/Project Manager Female 13 19 S1 Games Audio Freelancer (male) Male 13 20 S2 Games Lecturer (female) Female 8 21 S3 Employer and Industry

Representative (male) Male 12

Total 224.5 *Coding scheme outlined in Section 4.4.4; ** Approximate number of pages resulting from the interview, where pages are A4, 12point, single spaced (not including questionnaire responses); ***combined with M9 as both attended the same interview

This empirical data offered a factual description of the participants, including

basic demographic information regarding their cultural background, age and number of

children. As Table 28 illustrates, a notable point is that only two of the 18 women have

children. There are, however, limitations to the demographic data as not all participants

provided this information; it was voluntary to do so.

Table 28 Overview of female DCI participants

* Assumed to be Australian; however, participants were not required to provide this information. ** Several participants did not provide their age. *** One participant worked in a multimedia role within a games company, so identified with both sectors.

Findings indicate that the women employed in the DCI have diverse backgrounds

and pathways into the industry. Participants’ previous occupations included roles in

education and in publishing. A majority of the participants had TAFE62 or university

undergraduate qualifications, several had a Masters qualification, and one, a PhD degree.

Not all qualifications were related to the DCI. For example one had a degree majoring in

Russian. The majority of the female interactive content creators interviewed were in their

early career; this was defined as having less than 5 years of employment in the DCI

industry. For most participants, this was their first or second role in the DCI. Focusing on

the early career period provided access to a rich intersection of participants’ recollections

of their situated or immediate work place, their recent history (for example, entry into the

workforce or workplace), and of their future career considerations or intentions to remain

in the industry.

DESCRIPTION OF STAKEHOLDERS

As Table 29 illustrates, there were three stakeholders interviewed: 1) a male

freelancer, 2) a male employer and industry representative, and 3) a female educator and

industry representative. Stakeholders were sampled for maximum variation of the

insights they could offer. Stakeholder 1 worked as a freelancer in the games industry;

however, he also had qualifications in multimedia development. Stakeholder 2 was a

male employer and an Australian multimedia industry representative who had a long

involvement in the industry (over a decade). The third, a female university educator and

Australian games industry representative, provided insights into education and training

aspects. The stakeholders were members of key national DCI organisations including

AIMIA (Australian and Interactive Media Industry of Australia), and international

organisations including IGDA (International Games Developers Association).

62 Technical and Vocational Education qualifications range from level 1-6 (Certificates and Diplomas)

Age Industry Cultural background Children

** 20-25 25-30 >30

5 2 5 6 9*** 9 15* 1 1 1 16 2

Table 29 Overview of stakeholder participants Stakeholder type Industry sector

Freelancer Employer Educator Games Multimedia 1 1 (male) x x

2 1(female) x 3 1 (male) x

DESCRIPTION OF THE ORGANISATIONS

As Section 4.2.1 of Chapter 4 outlined, participants were recruited from both

games and multimedia organisations, with the researcher actively seeking an equal

number of participants from both. Participants were employed across 11 DCI

organisations, 10 of which were geographically located in Brisbane. The one

organisation not based in Brisbane was a large international organisation; however, the

participant employed at the international organisation had worked at several Brisbane

organisations prior to her current position. Games organisations ranged from start-ups

who had published independent titles, in contrast to larger multinationals that tended to

produce well-proven genres from established publishers. Multimedia organisations

ranged from small family run start-ups to larger organisations established for over a

decade. Table 30 summarises the organisations in which participants were employed;

however, certain information is omitted to maintain participant anonymity. Appendix 7

provides further information on several of the organisations.

Table 30 Case study organisations (including games and multimedia production) No. Organisation Code Type of organisation

1 Krome > B Australian

2 THQ Australia > H US parent

3 Creative Assembly > G UK parent

4 Pandemic > F US parent

5 * K** UK large multinational

6 Ice Media > A Australian

7 Queensland Transport D Australian (State Government)

8 Queensland Museum E Australian (State Government)

9 * C Australian (family run web)

10 ABC Online News > I Australian

11 BCM > J Australian

* Name not provided to ensure participant anonymity; ** An international organisation; however, the experiences of the participant included those while working in Brisbane, Australia in DCI roles. *** A fourth provided a questionnaire; however, the interview was not recorded due to a technical problem with equipment

DESCRIPTION OF THE OCCUPATIONAL ROLES

The roles the women were employed in varied, reflecting the exploratory nature

of the case study and the maximum variation sampling approach. Roles included:

artificial intelligence (AI) programmer, web designer, character artist, software engineer,

and social media coordinator (See Table 27 for the complete list). The participants self-

identified their roles, in either their questionnaire responses or during interview. These

roles fit within the ABS defined occupational class of ‘interactive content creator’.

However, the participants themselves were often not aware of this occupational

classification.

Participants identified that many DCI occupational roles were related. There was

a “blurring of lines between multimedia production and gaming as well and graphic

design. They’re all sort of interconnected so if you’re good at one you could easily land in

a job in another field that’s related to that” (S1). There was a difficulty in defining what

multimedia roles entailed, which can be attributed to the industry’s relative youth or

emerging status 5

63: “I think it’s because the industry itself is not mature enough” (M7);

and to the ambiguity surrounding the industry: “Any person you ask what multimedia is

will give you a different response” (M7). There may be scope for the DCI to convey the

occupational classification of ‘interactive content creator’ as a way of minimising the

ambiguity surrounding roles in the DCI.

DESCRIPTION OF WOMEN ’S UNDER-REPRESENTATION

Analysis of the empirical data (including interviews and secondary data) revealed

that women are under-represented as interactive content creators in the Australian DCI.

A male freelancer observed: “There’s definitely not equitable numbers” (S1). Under-

representation is most evident in games production organisations. For example, of the 80

staff employed in a Brisbane studio of an international games development organisation

(Organisation H in Table 30), only two were women. In another games organisation

(Organisation G), only two of the 50 employees were women. Similarly, there are

inequitable numbers in multimedia organisations, although there was an assumption that

there were more women entering multimedia organisations. “There are more and more

girls getting involved…there are definitely more males” (S2).

Secondary sources of data support the finding that women are under-represented.

For example, a 2009 online web article reported that, in Brisbane-based games

organisation THQ studio, only five of the 100 staff were women (Appendix 17, Item 49).

An international survey of games workers reported similarly low rates: “I'm a girl 63 The researcher experienced an interesting insight when interviewing one informant, who seven years earlier had been her student. The researcher and the interviewee both acknowledged that the interviewee’s occupational role had not existed 7 years earlier.

working in the games…where I work we have no other females working on games”

(Appendix 17, Item 28 -F, 31, White, Uni, UK).

Although women were under-represented, participation rates varied across

different contexts. For example, public sector organisations, in contrast to private, were

seen as being more equitable: “It's one thing I've noticed is that it is seems to be quite

equal of male and female” (M2). The different characteristics or types of DCI

organisation (for example, public and private) warrant further investigation.

Low participation rates also manifested in related educational contexts. One

stakeholder (interviewed in 2010) described there being a “phenomenal amount of

women” (S2) participating in a Brisbane university degree program related to games

production. The stakeholder attributed this ‘high’ number to the course content, which

was about design “concepts, arts, visualisation, and design” (S2), thus suggesting that it

was more appealing to females. However, this ‘phenomenal’ amount comprised only 20

female students out of 100: a ratio of 1 to 5. A similar pattern appears internationally,

where a respondent to the IGDA Diversity survey noted the low rates of female students:

I am attending university now in game development, out of 65 students, there are only three females. (Appendix 17, Item 28-M, 18, White, Uni, Australia)

Participation numbers varied over time. One participant (M7) described the

“ridiculous” situation where, although the gender ratio had been equitable in the year she

began studying a multimedia degree (2003), a few years later there was only one female

out of maybe 30 or 40 in their second or third year of study. Further investigation of the

patterns of women’s participation (over time) is warranted.

Participation numbers varied across different occupational roles. Although

participants identified that it was more common for women to be employed in design and

production roles and less in technical roles such as programming and Quality Assurance

(QA-related roles), women appeared to be under-represented in art roles: “I don’t know

one female concept artist. I know like 40 or 50 males” (G7); and “Although it's still a

majority of males working in the art roles and definitely design roles as well. There are

more girls working on art than in the other teams. And I think there is more girls working

in design than in programming” (G3). Secondary sources of data, such as the IGDA

report59F64 (IGDA, 2005a), also indicate that women’s participation varies across different

types of occupational roles: “Females in my experience tend to be in the more creative

areas of the project, like design, artist. I have not come across any female programmers

for some reason” (Appendix 17, Item 28-M, 23, White, Uni, Australia). Other

respondents in the IGDA survey state: “From my experience, less than 10 percent of a

64 Game Developer Demographics: An exploration of workforce diversity –Comments on Diversity Report

QA team for a development project will be female” (Appendix 17, Item 28-M, 37,

White, Uni, USA).Further investigation of women’s occupational segregation in the DC

is warranted.

DESCRIPTION OF INFLUENCES

Women in the DCI face a myriad of influences that can challenge or support their

participation. For example, the participants recounted their limited access to technology

in their childhood. Access to technology appears to be an underpinning resource as it

fosters skill development and confidence in using technology. In later years, at school,

participants described a lack of information regarding career options in the DCI. The lack

of information constrained their career choices as it limited opportunities for informed

decision regarding a career in the DCI. However, rather than reporting a litany of

influences, the findings draw on the three frameworks proposed in Chapter 3 as a way of

identifying and understanding influences.

Framework 1 offers three categories with which to analyse the data: 1)

environment 2) person, and 3) interaction between 1 and 2. Findings presented in Section

5.2 reveal that a range of characteristics (within these categories) can influence women’s

participation. Framework 2 provides several concepts from the Human Agency theories

in the HAMTS(as presented in Table 11), such as norms and self-efficacy, with which to

analyse the data. These concepts foster further understanding of influences. For example,

as Section 5.3 reveals, Bandura’s SCT draws attention to the importance of same gender

role models in scaffolding skill development. Framework 3 requires that the phenomenon

be considered using Bhaskar’s ‘three domains of reality’: 1) empirical, 2) actual, and 3)

mechanisms. The emphasis of Framework 3 is on the identification or abstraction of

mechanisms that manifest in the ‘real’ domain. As Section 5.4 reveals, there are several

mechanisms that influence women’s participation in the DCI.

5.2 FRAMEWORK 1: ANALYTICAL (E, P, I)

The analytical framework fosters an investigation into three categories of

influences: 1) Environment (E), 2) Person (P), and 3) the Interaction between the two (I).

Accordingly, the empirical data collected for the exploratory case study was categorised

into these E, P and I categories of Framework 1. The findings extend and refine the three

categories of the guiding analytical framework. This refinement reflects Stage 2 of

Danermark et al.’s (2002) model, which involves the resolution of various components,

aspects, and dimensions.

As Section 5.2.1 (below) presents, the refinement of the Environment (E)

category leads to the identification of four dimensions of the environment: 1) social, 2)

cultural, 3) mediated, and 4) resource. Within those dimensions, several properties

manifest at either macro or micro levels. The refinement of the Person (P) category in

Section 5.2.2 leads to the identification of one dimension of social identity, which in turn

is comprised of two major properties, gender and occupation, and two minor properties,

age and race. The term ‘dimension’ is chosen to describe these categories; this is

because the term ‘category’ may imply a bounded relationship, whereas ‘dimension’

reflects interconnectedness. As Section 5.2.3 presents, the researcher faced a challenge

in refining the category of Interaction (I) by solely using inductive reasoning. Drawing

on the ontological underpinnings of Critical Realism helped to stimulate further analysis

and led to a focus on events, specifically events manifesting from the interaction of the

environment and person.

5.2.1 THE ENVIRONMENT (E)The analysis of empirical data led to the

refinement of the Environment (E) category in

Framework 1. Influences could be categorised across

four key dimensions, which in turn could be further refined by identifying properties

within those dimensions. The environment category is further refined by identifying the

temporal context and different levels of the environment, including macro and micro.

FOUR DIMENSIONS OF THE ENVIRONMENT

Chapter 4 discussed the data analysis techniques, which included pattern

analaysis. Pattern analysis of the empirical data led to the identification of four sub-

categories or dimensions within the Environment (E) category of the initial guiding

framework: 1) social, 2) cultural, 3) mediated, and 4) resources (as illustrated in Figure

8). Each of these dimensions is introduced below and described in greater detail shortly.

Figure 9: Refined ‘Environment’ (E) category

1) The social dimension recognises that an individual may experience different

social contexts (including communities of practice, family, education and work), and

other people or social agents (such as parent, peers, and colleagues) within those

contexts. As will be explained further on, the social dimension recognises a temporal

aspect as the contexts can be positioned along a typical lifespan comprising of childhood,

early and late education, early career, and career progression.

2) The cultural dimension recognises a wide range of societal influences. These

are, for the most part, characteristics of the face-to-face environment a person

experiences. As will be further outlined, data analysis resulted in the cultural dimension

being conceptualised as encompassing: social beliefs and practices; and linguistic,

geographic, economic, and political properties.

3) The mediated dimension recognises that not all interactions between people

are face to face. Rather, they are mediated by other mediums of communication, such as

literature, television, mobile phone, and technologies such as the internet.

(4) The resources dimension recognises that objects or entities within the

environment may influence participation. As will be further described, the analysis of

empirical data reveals that the resources dimension includes information (for example,

learning resources), artefacts (for example, computer games) and technology (for

example, computers).

TEMPORAL ASPECTS OF THE ENVIRONMENT

Participants recounted influences as occurring in their childhood, during

schooling, within the current workplace context, and even as future possibilities. The

researcher interpreted that participants positioned their participation along five lifespan

stages (See Table 31), starting from birth through to adult years. At the childhood

lifespan stage (birth to early school), typical environments women encountered were the

family home and school. In later life, when women are in the workforce, typical

environments are workplaces. As Figure 10 indicates, the focus of the current case study

is the industry context, which aligns to the workforce life stage (Stage 4 in Table 31).

Table 31 Lifespan stages and contexts of participation Lifespan stages Typical environment 1 Childhood- birth to early schooling Family, school 2 Early education-junior to high school Family, school, hobby and sporting contexts 3 Late education-university or further education School 4 Workforce entry- early years of career entry Workplace 5 Workforce progression-career advancement Workplace

Figure 10: ‘Early career’ context within the social dimension of the ‘Environment’ (E) category

Participants’ accounts suggested that their early experiences informed later life

experiences. For example, childhood exposure to computers at “an early age” (G6) had an

influence later in life as it fostered their self-efficacy towards technology. Participants

noted that it was difficult to maintain their interest in technology during their adolescence

(Life Stage 2 in Table 31) if their female friends didn’t share the interest.

One participant reminisced how, at the age of 6, she had girlfriends with whom

she shared an interest in computers: “I had good girlfriends who also had Commodore

64's at home so we'd have commodore 64 parties and that was just a part of the group”

(G2). However, she noted this changed in her adolescence: “Then as I grow older I find I

don't have as much in common with girls as much as boys”. Another identified the skills

she developed as a teenager assisted her in her current role in the DCI. She had “spent a

lot of time like as a teenager online reading blogs and chatting [unclear] in forums” , and

this provided her with “quite a lot of experience” (M8), and led her to be “pretty good with

computers [omitted] compared to probably the average person” (M10). Thus, childhood

experiences had an effect on participation later in life. Table 32 presents examples of

influences categorised to both the four different dimensions of the environment and the

lifespan stages.

As discussed in the next chapter, organising the description in a manner that

keeps sight of the temporal aspects fosters more holistic findings by identifying the

influences that manifest at a certain life stage, and those that also surface across life

stages. Thus, temporal aspects must be recognised when seeking to understand

influences.

Early career lifespan stage

Table 32 Matrix of influences across the four dimensions of the ‘Environment’ (E) category from a life stage perspective

Dimensions Social Cultural Mediated Resources

Context Social agents

Childhood Family Parents, family members

Customs and practices

Television Technology; for example, computer

Early education

Educational institution

Teachers, peers Socialisation practices

Magazines Literature, technology

Late education

Peers, teachers Gender ratio Training material

Early career Workplace setting

Colleagues, employers, mentors

Policies Television Email, websites

Career progression

Conference setting

Family,employer, colleagues

Organisation culture

Access to child care

CoP Online gaming

Peers WoW 60F

65 Computers

MACRO AND M ICRO ASPECTS OF THE ENVIRONMENT

Further pattern analysis reveals that the four dimensions of the Environment (E)

may be further refined by identifying characteristics of the macro level environment (for

example, global or national setting), and the micro level (for example, organisation

setting). Participants did not explicitly identify ‘macro’ or ‘micro’ influences; rather, this

categorisation process was achieved by the researcher’s interpretation of the data.

As Figure 11 illustrates, once the Lifespan Stage (A) (for example, early career)

has been recognised, the social, cultural, mediating and resource dimensions (B) can be

explored at both the micro (C) and macro (D) level. The four dimensions of the

environment (B) can be used to explore the macro level characteristics (D). Participants’

comments suggested that, at the macro level, the cultural dimension includes properties

such as legal –national EEO legislation, national policy – AWA’s61F

66, and economic –

dot.com crashes and the GFC (Global Financial Crisis)62

67. The mediated dimension

includes mainstream television, radio, the internet and newspapers. The resources

dimension includes technological infrastructure; for example, access to affordable and

reliable internet or broadband services.

Similarly, the four dimensions of the environment may be used to explore the

micro level influences (C). For example, participants’ comments suggested that: at the

micro level, the social dimension involves people in the workplace; the cultural

dimension involves the characteristics of the organisational culture such as policies; the

mediated dimension involves the localised effects of media such as industry newsletters 65 World of Warcraft (WoW) is a massively multiplayer online role-playing game (MMORPG). 66 Australian Workplace Agreements (AWAs) were proposed by the Australian government in the mid-2000s. 67 Participants indicated they were aware of ‘legislation’ and this may refer to specific legislation, such as Equal Opportunity Act 1995, and Equal Opportunity for Women in the Workplace Act 1999; however, these were not explicitly mentioned.

and websites; and the resource dimension involves the resources available to the

individual within the organisation such as computers. Thus, influences can manifest at

both the macro and micro levels of the environment.

Figure 11: Macro and micro levels across the four dimensions of the Environment (E)

There were links between macro and micro influences. For example, several

participants noted that the general economic health of the country (macro level

characteristic) could influence local hiring practices (micro level characteristic): “Games

industry itself right now is suffering from the whole economic drama worldwide and we’ve

had several large companies just locally be closed down and a whole bunch of people

were out of work” (S1).

Secondary sources suggest outsourcing may further reduce local opportunities

(See Appendix 17, Item 38: an advertisement for overseas labour for as little as $4 an

hour). If international games companies outsource to cheaper labour in countries other

than Australia, this would ultimately influence an individual’s micro environment by

constraining their salary potential and employment opportunities.

Identifying a relationship between macro and micro level influences helps view

findings in a relational and unified manner. Table 33 identifies the properties manifesting

at either the macro and micro levels across the four dimensions of the environment, and

provides examples of the influences within those properties. Findings indicate that

although each of the four dimensions manifest across the macro and micro levels,

properties may or may not emerge at both. For example, the ‘economic’ property

manifested at both macro and micro levels. There was evidence of global economic

properties such as the dot.com crash and the Global Financial Crisis (GFC) at the macro

level, and local hiring practices at the micro level. However, the language property

manifests only at the local (micro) level in the form of jargon (relevant to technology and

National legislation

Mainstream television

National broadband infrastructure

Organisational policies

Localised media

Organisational resources for example computer

software). Further research of properties across both macro and micro levels of the

environment is warranted.

PROPERTIES OF THE ENVIRONMENT

As Table 33 shows the four dimensions of the environment were further refined

to identify properties within those dimensions. These properties include: the social agents

in the social dimension; the gender ratio and workplace practices in the cultural

dimension; the mediums of the internet in the mediated dimension; and resources (such

as magazines) in the resources dimension. The identification of properties within each

dimension further organised the rich description. The following section provides further

descriptive insights of several properties within each of the four dimensions.

Table 33 Examples of influence in the DCI within the environment dimension Dimensions/Property Macro level influence Micro level influence Social Agents

Family Father, mother, brother, child Peers Male colleagues in the workplace

Role models Few role models Mentors Often male

Settings a) Family Home

b) Education Educational institution Specific student cohort c) Work Industry associations, parent

company Social groups in the workplace; for example, sporting groups, industry associations

d) Communities of practice

WOW, conferences Special Interest Group (SIGs)

Cultural Physical Open plan offices Ratio Gender ratio approx. 50% Gender ratio approx. 11% in games

organisations Language Technical jargon Economic Dot.com (2001), Global Financial

Crises (2009), job insecurity Local redundancy

Customs/practices Gaming becoming more mainstream and family friendly

Garage versus professional practices: beer and pizza culture, casual clothing

Legislation and Policy AWA work agreements; Fair Work Act 2010

Unclear organisational policies; no union or occupational awards

Historical Women’s right to work Few women in the pipeline Geographic Australia’s geographic isolation Brisbane a games hub Mediated Medium (television, magazines, internet)

Lack of roles models in TV programs

Presenters at industry conferences predominantly male

Resources People Lack of female mentors Information Books, internet Specific websites, for example,

Gamasutra Technology Affordability and availability of

computers SDKs, software, file exchange facilities

1) SOCIAL DIMENSION

Recognition of the social dimension fosters a description of the 1)

social setting, and 2) of the social agents or people (within the setting).

Although the case study investigates women’s participation in DCI

organisations, participants often described their current participation in

relation to their past experiences. For this reason, the conceptualisation of the

social dimension recognises a lifespan perspective and identifies various social

settings that a person encounters in their life (as identified in Table 31).

Table 34 Micro context related to lifespan stages Lifespan stages Social setting 1 Childhood-birth to early schooling Home

Communities of Practice (CoPs); (for example, social groups)

2 Early education-junior to high school Education

3 Late education-university or further education 4 Workforce entry- early years of career entry

Work 5 Workforce progression-career advancement

There are four social settings 1) home, 2) education, 3) work and 4) CoP, as

presented in Table 35. Participants identified home setting most often when talking about

their childhood), education (schooling period) and work (career period). The

communities of practice (CoP) may occur at any point of the lifespan and are

exemplified by social networking phenomena such as Facebook.68,

69 Many games

companies maintain social media profiles via Facebook (See example in Appendix 17,

Item 5). In addition, these settings include the people (social agents) within those social

settings. Participants identified social agents as including family, friends, teachers,

university professors, and peers, as illustrated in Table 35.

Table 35 Social agents identified within social settings

1) Social Settings 2) Social Agents

Family Friends Peers/ colleagues

Teachers Employer Role models/ /Mentors

a) Home x b) Education x x x x c) Work x x x x d) CoP x

X indicates clear evidence (from empirical data) of the agent within the micro context of the social setting

A) HOME

Participants described characteristics of their home setting as influencing their

participation. Most made reference to home settings when reflecting on their childhood,

although a few noted their adult home settings as well. Key social agents in their

childhood were family members including father, siblings and to a lesser extent their

68 The definition of a CoP follows Wenger (1998). 69 A popular online social networking website site: http://www.facebook.com

mothers. Fathers were often mentioned as being the facilitators of access to technology:

“I was quite young, like probably 9 or 10 he would show me like things on the internet.

Yeah, so he introduced me to it at a very young age” (M8). Other participants recounted

positive memories of playing computer games with their brothers: “My brothers were into

it…I’ve played games all my life” (M10). She recounted how her brother had facilitated

her early access to technology: “My brother did an IT degree and he brought home an XT

that he built at uni and very proudly gave it to the family and let us play with it” (M10).

One participant, an Artificial Intelligence (AI) programmer, commented on how

supportive both her father and mother had been, consequently minimising the influence

of the constraining stereotypes surrounding women’s capacity to use technology: “Well

right from when I was very little I never like had a negative stereotype of getting into that

kind of thing [computing]. Whereas most women would think oh no, that’s not me” (G3).

Another participant suggested that “if you’re involved with that from a very early age you

grow up thinking 'oh it's not just a guy thing' it's something that girls can do” (G5). Thus,

family members were seen as an influence on women’s participation.

B) EDUCATION

Participants described characteristics of their education setting as influencing

their participation. Education settings included TAFE, short courses, and university. Key

agents in these settings were teachers and fellow students. As noted earlier, participants

would often be in the gendered minority in these education settings: “All the subjects I

have chosen like math’s and animation it’s predominantly a male orientated environment,

like I’ve probably been 1 of 3 girls in anything I’ve tried" (M2). Participates also noted that

education settings did not always provide the conditions for developing skills suitable for

a DCI career:

The programming we did at high school was completely different it was really outdated

on Macintosh computers…nothing to do with any sort of programming at the end of the

day and it didn't really give an accurate representation...no one liked the teacher who

taught it ...I think most people’s only access to it was for a few weeks in Grade 8 in this

sort of negative environment. (G3)

C) WORK

Participants described the characteristics of their work setting that had influenced

their participation. Work settings included a range of organisation types such as start-up,

multinational, private, and public. Key agents in these contexts were mostly male

colleagues, employers, clients, and publishers. As the case study focuses primarily on

women’s career participation, further characteristics of DCI organisations are detailed

shortly.

D) COMMUNITIES OF PRACTICE (COP)

Participants described the characteristics of several social and professional

groups to which they belonged. These groups included social networking groups on

Facebook, online gaming groups [such as World of Warcraft (WoW)], sporting groups,

and industry-related functions such as conferences. The key characteristic of these

groups was that they provided a sense of community to the individual. Thus, this setting

is conceptualised as a Community of Practice (CoP). The CoP could run in parallel to the

home, education and work settings. The CoP often fostered skill development related to

the technology employed in the DCI, such as computer and the internet skills. For

example, a participant noted that although her school provided very little in the way of

computer resources, she gained experience by participating in online communities, where

she learnt basic programming skills. Participant M3 described how, through playing

games online, she developed an interest in modding (modifying aspects of the game) and

thus extended her technical skills. Key agents in these settings varied greatly, being

dependant on the nature of the community, but included fellow gamers and other

industry professionals. Further research could consider the relationships between the

CoP and other settings.

2) CULTURAL DIMENSION

Cultural characteristics of the environment, specifically the

micro context of the DCI organisational context, are an influence on

women’s participation. DCI organisations may present an uninviting

workplace culture to women. A male games freelancer described the

influence of the primarily “really, really male” culture as a “bleeding

obvious” (S1). He suggested that women might not find the male-

orientated culture that existed within DCI organisations as appealing.

The male freelancer conveyed that he did not personally relate to the ‘male’ culture;

however, because he was a male he could more readily position himself within it.

Secondary sources of data indicate a similar situation in the related IT Industry

(Appendix 17, Item 47):

I would probably socially know how to handle people like that a little better and know

where to place myself to get by and to deal with them and to make it work. But I could

see how it would be uncomfortable potentially for women. (S1)

Several participants explicitly mentioned the influence of organisational culture.

There were cultural differences between different organisations. For example, the

multinational’s “professional” (M2) and “corporate culture” (G4) contrasted with the

independents that “started off as a couple of guys and all that sort of thing” (G4).

Government organisations were contrasted to private organisations as having a “different

culture” (M2). DCI organisations also appear to have sub-cultures, formed through shared

interests among colleagues:

There’s a lot people who love pirates and it seems silly but…people ...share common

interests…I think working in a specialised field, you tend to get people who are quite

similar and yeh we all seem to get along quite well. So those would be advantages. I

think there's a nice culture here. (G4)

Analysis of the empirical data led to the dimension of culture being further

refined to identify several specific properties including: a) physical setting, b) ratio, c)

language, d) economics, e) customs and practices, f) legislation and policy, and g)

historical properties. As the case study is concerned with women’s participation in the

DCI, the evidence below mostly pertains to the properties that manifest within the micro

context of the organisation.

2A) PHYSICAL PROPERTY

Participants referred to the size of the organisation (with reference to the

number of employees) as being a possible influence. There was a preference for

medium sized organisations:

You're looking at a company that's got 50 people or less to get that quality of

workplace conditions which is far more rewarding than working for a larger

company. (M5)

Small organisations could create a supportive environment because

everyone knew each other: “Like you know everyone’s name and you can have a

chat with everyone” (M1). However, the negative aspect of a small organisation

was that employees could feel as though they were micro managed and that they

could not avoid certain colleagues: “There are less people so the interaction is

harder, and it’s harder to avoid people” (M1). Organisations needed to be “small

enough” to “keep personal touch with people” (M1). Large organisations were also

seen as being linked to poor life/work balance:

Basically you can't have life if you’re working in advertising agencies, especially

in the big network. So I thought no I don't want to work in that kind of

environment I would prefer a little design studio. (M1)

Additionally, participants suggested that the physical layout of their work

environment could be an influence. Open plan spaces facilitated greater

communication between workers: “It was open plan so we were working together

all the time ...” (G3). Field observations confirmed that the organisations B, H, and

A (See Table 30) had open plan layouts for workers; however, other organisations

were not directly observed. Thus, open plan medium-sized organisations can foster

2B) RATIO PROPERTY

Findings suggest that it is not only the number of women in an organisation

that matters; it may be the ratio of women that influences their participation. For

example, one of the largest games companies had the highest number of women

employed: of the 280 employees, approximately 28 were women. However, its

rapid growth had led to the isolation of the female staff. The environment had

become “so big all the girls had sort of been scattered around” and they might not

“even pass each other in the corridor” (G1). The gender ratio influences the sense

of belonging women may feel, with lower ratios making women feel as though

they were in the “minority” (G1). However, as one employer suggested, although

the ratios create unbalanced work environments, they were difficult to address.

I think games [presents an] unbalanced work environment... But it's a fact so...to

address that balance is pretty difficult for them to do. You're swimming against

the tide for the one or two women that do sort of hang out there. (S3)

In addition to the gender ratios in a workplace, there was a need to consider

the ratio of occupational roles. One participant described feeling professionally

isolated, not because she was the only female in her workplace, but moreso

because she was the “only designer” (M6). She expressed the need to interact with

other designers so that she could develop her skills.

Although women were in the minority in the DCI, participants’ accounts

suggested they could ‘cope’ with being in the minority; however, this “takes a

while to get used to” (G8). As one participant noted, attending a meeting where

“there's 30 guys there and you're the only female” can be “a little bit weird” ;

however ,“I don’t know if that's a barrier or not”, although it would be “better if

there were some more females” (G8). Thus, the evidence suggests that although

gender and occupational ratios are an influence, they may not be a barrier to

2C) LANGUAGE PROPERTY

Language manifested as a property of the environment. The DCI appears to

value workers with specialist technical knowledge of software and hardware, and

this knowledge often involved technical jargon. Participants identified the need to

learn technical jargon as it fostered a sense of confidence and belonging to the

work environment. One participant recounted that “the jargon can be quite

intimidating but once you know it, it's really simple to understand” (G1). Another

participant was surprised she picked up the jargon so quickly: “I picked up you

know the jargon really quickly, knew what was going on, knew what I was looking

for, it all seemed to come really naturally, I'm like ohh” (G1). Thus, the evidence

suggests that mastery of language (as a part of culture) is an influence on women’s

participation.

2D) ECONOMIC PROPERTY

The property of economy influences women’s participation. Earlier

findings identified that the economic property could manifest at the macro level in

the form of global financial conditions. Micro level economic properties included

low levels of remuneration for DCI workers. Participants suggested other

industries offer higher salaries; thus, transitioning from other industries could

involve a pay cut: “She'd be going from sort of a 80-90 thousand job to entry level

position in the company which I’m sure she'd be willing to do but it's very difficult

to” (G2).

Participants suggested workers may leave the DCI because they realise

they can be better remunerated in other industries. For example, in “admin

assistant roles you can use your basic multimedia skills to do up the brochures and

things like that, so I'm sure they utilise their skills elsewhere and there probably is

more money in some of those roles than there is when they start out within the

private sector” (M2). It was suggested that low salaries manifested as a result of

people wanting to get an initial start in their careers. There are people who “will

work for nothing" (G7). Lower salaries may also correlate with the participants

being in their early career stage. There was a high level of ambiguity in regards to

what remuneration might be expected: “In this industry salary is a little

questionable” (G8). For example, women in games roles identified that ‘games’

paid low salaries. In contrast, the women in multimedia saw games roles as being

lucrative: “I would think the gaming industry in particular is quite well paid although

you do horrendous hours” (M2). These were similar findings to the Insight

Economics 2006 Australian Electronic Game Industry Report (Appendix 17, Item

54) which found that in a survey of games workers, there was considerable

variation in salaries for some occupations. Thus, the evidence suggests that there is

ambiguity surrounding salaries in the DCI.

There was often a trade off between salary and working conditions. The

more a person earns the less likely they would have a life/work balance: “If you

work for EA in Canada you'll definitely make 6 figures if you’re good and umm if

you work for a smaller company you make a lot less but you get a way better work

environment so it's all a trade off but definitely games pays less. I think that

perception is within the industry” (G3). One participant noted that the 2005 EA

spouse case (See Appendix 17, Item 34) in the US had fostered “better conditions”

for the industry: “So I think every now and then there’s a big kick in the pants for

the industry … so now there are a lot better conditions” (G2). Thus, the evidence

suggests that women in the industry perceive that the higher the salary DCI

workers received, the less likely they would have a life/work balance. In general,

women valued both salary and life/work balance.

2E) CUSTOMS AND PRACTICES PROPERTY

Participants identified a diverse range of customs and practices that can

influence women’s participation. A widely-recognised industry practice was that

of working long hours. Secondary sources of data confirm that unappealing

working conditions are a characteristic of the games industry. For example, a

keynote presentation at the 2004 Australian Gaming Developers Conference

discussed staff morale, crunch time and unpaid overtime, and identified that the

Australia’s games industry was “characterised by many of the poor working

conditions being debated overseas” (Appendix 17, Item 37).66F

70 Thus, the evidence

suggests that there is a perception that working in the DCI entails long work hours.

However, participants in the current case study noted their hours of work

were mostly reasonable. Table 36 summarises the hours worked per week by

participants interviewed in Phase 1. One participant suggested that her previous

position as a schoolteacher had required her to work longer hours than she did in

the games industry. Thus, the notion that working in the DCI entails long working

hours can be challenged by the findings.

Table 36 Hours worked by Phase 1 participants ID Organisation

Role Time in

role (months)

Time in industry (months) *

Average hours/wk

M1 Multimedia Creative Director 20 20 42

M2 Multimedia Multimedia Developer 30 70 40

M3 Multimedia Software Engineer 14 14 40

M4 Multimedia New Media Producer 16 60 45

M5 Multimedia Web Developer 6 24 40

M6 Multimedia Web Developer/ Designer

11 18 30**

G1 Games Production Assistant 18 18 38

G2 Games Junior Props Artist 18 44 50

G3 Games Artificial Intelligence Programmer

8 54 40

G4 Games Project Coordinator 6 24 40

G5 Games Assistant Producer 11 24 40

G6 Games Games designer 26 26 50 *This did not include any volunteer work (for example, time volunteered when a student, or personal time spent developing skills or a portfolio). ** This participant was the only one in the study who worked part-time.

70 The ‘spouse’ in the EA case now provides a website that reports on the conditions of work in games organisations: http://www.gamewatch.org/phpbb3/viewforum.php?f=6.

The practice of working long hours differed between public and private

organisations. Private organisations demanded longer work hours, and public

organisations were seen as offering greater flexibility: I’m assuming anything that’s

private has no work-life balance, straightaway” (M7); and“ Yeh I think in the private

industry I think you’re going to have to dedicate more than just 9-5 hours, not only

just for your job…but also the out of work hours” (M2).

The two women interviewed from the public sector put forward that the

main attraction of working in public organisations was the life/work balance they

offered: “Work life balance is quite important in the government” (M2); and“The

people are fantastic, very, very flexible and when you think about the future and

thinking when I have a family, there’s no way in the world [you’ll] work in the

private sector, no way in the world” (M7). For those participants working in private

organisations, there were few role models who had achieved the life/work balance:

“I haven’t really seen anyone in the games industry perfect the balance” (G2).

Although several participants suggested that public sector organisations offered

better life/work balance, one participant working in the public sector (who had a

child) identified that there remained difficulties in managing family and work

responsibilities: “I'd like a balance, I don't know, it's kinda hard, sometimes I get

really frustrated” (M4).

The practice of working long hours manifested for several different

reasons. These included poor management practices, demands of overseas

publishers, and unrealistic production deadlines: “In the beginning we used to work

insane weeks I mean 60-80 hours just to get a product through...A lot of it had to

do with poor management” (M4); and “My manager basically promised certain

deadlines…I think the management – it was a fantastic team but the management

practices were non-existent in terms of understanding project management” (M7).

These poor management skills were to some extent to do with the newness of the

industry: “The way these companies are managed or not managed is not the same

as many other more established industries” (S1).

The actual hours worked differed across different roles. For example, those

in Quality Assurance may work longer hours than concept artists: “There are some

people that aren't fortunate, like their role kind of mandates that they have to stay

late, like QA, they'll have to wait for a build or they'll have to hang around to

upload” (G4). Other roles within the same organisation would not face these work

demands:

So like my partner xxx was working on xxx as a concept artist so he got to go

home every day at 5 o’clock, 5.30...whereas my mate who's working on the

same project as an animator he could be there to 1 or two o’clock in the

morning…it is role specific, production type roles are quite easy 9-5. (G1)

In addition, there was a general perception amongst participants that

programmers worked the longest hours. However, the experience of the

programmers themselves was that they felt they worked reasonable hours. The

game programmer noted that the hours she worked were “pretty comparable to

most of my friends who work in various industries” (G3). A programmer in a

multimedia organisation made a point that she had only been asked to work back a

handful of times in the last year and, in each case, she was paid overtime. Thus, a

blanket statement cannot be made to suggest that all roles in the DCI entail

working long hours.

The practice of working long hours was dependant on structural

characteristics of the industry, such as the project production cycles. Cycles could

run for years; for example: “The Nickelodeon ones they have eighteen month

cycles” (G4). Long hours tended to manifest around “crunch time” (G4); that is,

towards the end of a cycle when a product was required to be completed by a

particular deadline. The long hours were also dependant on the projects

themselves: “The last few projects we worked public holidays and we worked a

Wednesday evening and then sometimes a Thursday evening and it was

expected, whereas now it’s like if you don’t sort of have the work you don’t have to

stay” (G7). If projects entailed participants learning new skills then this contributed

to the long hours they worked. If participants could use existing skills, there was

less of an expectation to upskill or “go home and do a lot of research or do a lot of

learning stuff” (G7). A multimedia programmer noted that when she was required

to learn a new skill–for example, a programming language–she was given extra

time in her work schedule. Thus, the evidence suggests that the practice of working

long hours can manifest differently depending on management approaches.

The practice of working long hours was not always an outcome of the

environmental characteristics, but could emerge from the individual’s actions: “I

think it's also coming from the individuals who establish that, not just the company.

Because there are people who literally work themselves to death and we probably

can’t do anything about that. I think people set their own boundaries” (G4). Some

individuals worked long hours because they were invested in and ‘loved’ their job.

Others did so in fear of losing their job; they felt they were “kind of expendable as

there are a lot of people looking to come into the games industry at the lower end”

One participant identified that although “every job” she had involved long

hours, she had contributed to these long hours: “I guess it’s also up to the person. If

you like to tick everything off for the next day […] sometimes you have to invest a

little bit more” (G5). Another participant admitted: “I’m sure people can do it in

shorter hours, maybe it’s - I think sometimes it can be your personality. My

personality, I can’t let that go until I know it’s okay” (M9). Yet another (M7)

suggested that it is an individual’s drive to produce quality work rather than “bad

products” that sustained the practice of long hours. A stakeholder described the

pressure to work long hours when aiming to develop quality products:

Like you’ve got school leavers who are ambitious and want to prove that they’re

really, really into this but there are others – a whole wave of them waiting next

year. So this ambitious student goes in, gives their all, gets completely burned

out by the time they’re 26 and gets spat out the other end and has to find a job

that’s actually going to just pay the bills and give them room for a life. (S1)

The influence of long hours varied over time due to participants’

circumstances. For example, although some described enjoying the long hours

when they first began their careers, this changed over time: “I got to the point

where I'll never do that ever again” (G8). For others, family responsibilities

conflicted with the demands of working life. Even when they did not have

children, these participants identified that working long hours would conflict with

such responsibilities: “I don't see how I could have been a mum at that time” (G8).

Thus, the evidence suggests that the practice of working long hours varies among

individuals.

Cultural practices manifested in other ways than the practice of working

long hours, including the clothing people wore and the food they ate. For example,

something as simple as the everyday clothing worn by workers appeared to be

important. Thus, the influence of clothing is later discussed. Even the practice or

custom of what foods one might be expected to eat when working in the DCI was

raised in participants’ accounts. It was suggested that there is a “beer and pizza”

culture that both women and men may encounter: “So if you’re not a beer and

pizza sort of person, it doesn’t matter whether you’re male or female” (G7). A job

advertisement for a games organisation suggests that eating healthy fruit is not the

‘norm’ in the industry context (Appendix 17, Item 51). Thus, findings reveal that

everyday cultural practices such as dress, food and work practices can influence

2F) LEGISLATIVE AND POLICY PROPERTY

A participant in a public sector organisation emphasised that clear policies

could support her participation. Section 5.2.3 further describes another

participant’s concerns about policies such as the Australian Workers Agreement

(AWA). Although participants mentioned policies, they made little reference to

macro level legislative characteristics that secondary sources draw attention to. For

example, there was no mention of The Equal Opportunity for Women in the

Workplace Act (1999), which aims to address social inequity across the wider

community. Further data and investigation of the influence of policy and

legislation is warranted.

2G) HISTORICAL PROPERTY

Findings indicate that cultural influences on women’s participation

manifest as a result of historical properties, such as society’s changing attitudes,

technological innovation, and specific historical events. For example, society’s

attitudes towards playing games have changed over the last few decades. As one

stakeholder noted, a decade or two ago games were “demonised” (S3), much like

the novel was when it first appeared. Conversely, another participant noted that

the games market had recently widened to encompass a broader target audience:

“Nintendo Wii is responsible for a lot of – like created an entirely new gaming

market, like retirement homes have the Nintendo” (S1). Thus, digital content

products may now be more appealing to a more diverse range of people than they

were in the past.

Access to technology has also changed over the participants’ lifetime.

Access to technology, such as computers, during their youth was integral to

women’s overall participation as it provided an opportunity for skill development.

However, women faced challenges in accessing computers, as they were simply

not readily available during their childhood and teenage years:

Whenever that was, 1994 or 1995, like there probably wouldn't have been a lot

of people necessarily who even had their own like computers in their houses or if

they had a computer they probably wouldn't have had the internet. (M8)

One stakeholder attributed the under-representation of women in the

industry to the reason that, historically, men formed the majority of technology

users. The same stakeholder revealed that in the last two years, only 2 of the 20

Flash developers in his organisation were women.

Interactive media, the websites, came very much from a programming kinda

base and mostly that's an IT related, male dominated sort of domain” thus “

there's a natural bias towards males in the industry because of the nature of the

way the industry has evolved…I don’t think there’s any reason why there

shouldn’t be women. I think it's a historical reason. I think it takes time for these

things to change through generations…Where typically in the 70's and 80's, you

know, when computers were evolving it was a male dominated area. So, it's not

surprising that there aren't more women; I'm not surprised that there aren't a lot

of women in the industry. (S3)

Another stakeholder also attributed the low number of women to historical

events; but reflected that although they had “absolutely no idea” why under-

representation occurs in the DCI context, it was for the “same reasons that girls

haven’t been engaging with IT for forever” (S2). As the comment below suggests,

secondary data challenges the empirical data. Where stakeholders suggested

historical reasons for women’s under-representation, secondary sources suggest

that the newness of the technology related industries should somehow negate the

historical biases and inequitable practices:

It's a relatively new field and people come in at relatively [the] same education level, [...] there has been a historical bias in many older fields reflecting past attitudes that men worked to support a family and women worked for pin money […] In IT, you may have started with a level playing field. (Appendix 14, Item 27)

Games organisations were perceived as having an emerging status, and this

lead to immature work practices: “This industry is still going through its

adolescence I think. It's not mature enough” (S1). The recent introduction of DCI-

related degrees at universities reflects this emerging status: “So this is just starting

to happen in a way. All of these game design courses are only just coming online”

(S2). The games industry is typified as arising from ‘garage’ style practices:

I mean it’s funny we as a company or as an industry spend a lot of money on

trying to understand organisational stuff and because I come from an industry

where that’s already defined, it’s pretty clear cut…in the games industry they’re

sort of muddling along so they haven’t really got that sorted out; that’s the age of

the industry. (G7)

In contrast, participants perceived multimedia organisations as being more

mature than the games organisations. Multimedia organisations were more

appealing because they strongly aligned to established industries (those with a

history and established work practices) such as publishing and advertising. It was a

“really integrated industry” (M9). Thus the differences between the multimedia and

games organisations warrant further investigation:

The interactive industry these days is more aligned with the public relations or

the advertising industry or the just think of social networking and social

marketers...so the multimedia industry or the interactive industry that AIMIA

represents certainly is a much more modern organisation with broad

communication objectives than the internet industry where it was all sort of ISP's

working out of garages making websites for people... (S3)

Historical events, including global economic events, could influence

women’s participation. Participants identified that both the dot.com crash of

(2000) and the Global Financial Crisis (GFC) could have influenced women’s

participation. It was perceived that these events lead to limited career opportunities

and lower salaries: “This is also about times because times have changed and

people are paid a lot more conservatively in the games industry than they were 10

years ago” (G7). Several participants mentioned that the influence of the 2001

dot.com tech bubble manifested well after the event: “Tech bubble…so the terrain

had changed and, yeah, the web design companies weren’t really doing that much

hiring. I was certainly looking for web jobs and I didn't get offered anything” (G8).

The 2009 global financial crisis (GFC) occurred during data collection and

may have contributed to the closure of one Brisbane games organisation and the

redundancy of approximately 80 people. Participants from another organisation

noted that the global event would lead to a lower number of local employment

opportunities. The “amount of like students being able to get jobs in games is

going to really go down” because “companies are not viable and positions simply

are not there” (G7). The participant suggested it would remain this way for 18

months to 2 years (they were interviewed in 2009). That historical economic event

would affect women’s and men’s potential for participation for a number of years.

As an addendum, three of the four Brisbane-based games organisations in this case

study faced major staff redundancies or closure in 2012. Thus, the evidence

suggests that characteristics such as economy and history can interact to influence

3) MEDIATED DIMENSION

The mediated dimension recognises the influences that manifest

through the media, in contrast to the in-person interactions within the

social dimension. A stakeholder, with several years of experience in the

industry, suggested the influence of the media “surfaces time and time

again” (S2). The media includes properties such as mediums including

television, literature, magazines, and digital products such as games.

These media convey social messages, such as norms. Table 37 lists

several of the mediums and messages identified by participants, and those found in

secondary sources of data. These messages include the messages that women do not

work in the industry, and that the industry is associated with males. The following

participant’s comment highlights that the media influences through the messages it

conveys:

There’s been a real amazing amount of backlash against the kind of independence of

women in the workforce I think over a period of time. We're talking about subtleness

which appears on TV; in you know interviews with people, in movies, magazines and

all of that sort of stuff and it all kinda progressively adds up and congeals on the

surface of what people perceive as the IT industry and women’s success in it. (M5)

Table 37 Examples of the types of media and messages within the mediated dimension Medium Example of message Evidence

Television Lack of female role models and appropriate images of women; stereotypical images of gender roles

G10, G5, G3

Industry magazines and events

Sexualised images of women appeal to males Appendix 17: Items 7, 21, and 25

Internet Advertisements for porn when accessing software sites

Appendix 17, Item 20

Games Sexualised imagery of women appeal to male audience

Promotional material

Conference featuring male majority

M1, Appendix 17, Item 6

Media can convey gender stereotypes. Secondary sources indicate that industry

initiatives, such as the 2011 Telstra FITT International Women's Day (IWD) Luncheon,

recognised that “women are still either stereotyped or ignored in the media” (Appendix

17, Item 55). Furthermore, the negative image of those involved with playing games is

“exacerbated by all sorts of media events” (S2).

Of the various mediums, television was mentioned by several participants. One

noted her concern with the medium’s inaccurate portrayal of women’s capacities and

enjoyment of working in the industry: “I'm just utterly shocked just in the advertising and

in the shows, the gender stereotypes that are still on there, now. [...] it doesn't fit in my

day to day life. The kinds of things I see women doing and men doing doesn’t really fit

into that stereotype I see on TV” (G3). Women were not only misrepresented but

television provided very few role models to women in the Australian DCI over the last

decade. A review of secondary data sources indicated that, in 2006, a television program

dedicated to reviewing games products was introduced on the ABC in Australia (See

Appendix 17, Item 39).71 However, the first female presenter was introduced in 2009,

three years after the program first aired.

Women were also under-represented in the internet-based medium of websites.

For example, a time-based analysis of the front page of a games industry website, 71 The website accompanying the television program has a backlog of broadcasted episodes from 2006 to the present.

Gamasutra, indicates that women’s presence changes from minimal in 2001, to stronger

in 2008 (See Appendix 17, Item 29). Similarly, resources available on the internet (such

as pirated software) appear to target males. Following up on a participant’s comment

about using a particular software package, ‘Dreamweaver’, the researcher conducted a

search in 2007 for this software on the internet. The search results led to several websites

that listed the software alongside links to pornographic images of women and live sex

shows (See Appendix 17, Item 20). Thus, this internet-based resource may be sending a

message that industry-related software is used primarily by men. However, a similar

search in 2009 revealed that sites where the same software could be downloaded featured

pictures of cute puppies (Appendix 17, Item 56); such images are less sexist.

Women were under-represented at industry conferences. One participant

identified that the majority of speakers at industry conferences were male: “For example,

every year I go to two big design conferences. One is in Sydney, Semi-permanent, and

one is in IGDA in Melbourne and its funny most of the speakers are men” (M1). Women

were also under-represented in the associated promotional material for these conferences.

An interviewee in an article featuring women in web roles noted the low rates of

participation of women at conferences: “With the exception of a women's blogging expo,

the number hovered at only around 15 per cent. The results are largely anecdotal and

have been contested by some of the conference organisers, but it does paint a fairly

gloomy picture” (Appendix 17, Item 52). Secondary sources, such as flyers and websites,

also indicate that men form the majority of speakers at international games conferences

(See Appendix 17, Item 6).

In addition, participants noted that media such as magazines that are typically

associated with women–for example, Cosmopolitan–did not promote the opportunities

available in the DCI: “Whether it’s from movies or TV or things in magazines, like when

you keep reading Cosmo or whatever and I never saw any articles…They’d never in a

million years have it on the radar that programming… Like it’s an amazing career. They

wouldn’t really know that” (M9). Furthermore, women are under-represented in industry-

related magazines. A random review of one issue of a key industry magazine (in the UK

and Australia) revealed that a majority of the industry members featured in the articles

are male (See Appendix 17, Item 30). For example, of the nine featured creatives, only

one is a woman. In the same magazine, an article features six authors, all of whom are

male. Similarly, the judges of a featured design competition are all male. Thus, the

evidence suggests that the mediums of television, internet and promotional material can

influence women’s participation.

4) RESOURCE DIMENSION

The resource dimension recognises that access to resources such

as technology, information, and networks (with other people) is an

influence on women’s participation. For example, having access to

technology such as computers from an early age was an important

influence, particularly when computers were not commonly available in

the wider community at the time most participants in the study were

children (between the 1970s and early 1990s). Family members, in

particular fathers, could facilitate access to these resources: “My dad’s a programmer so

we always had a computer around” (G3); and “I had the computer there. I just got

interested” (M7). Access to other resources, such as magazines and books, also fostered

skill development. One participant described reading a book on programming before she

even had a computer. As is the case for the previous three dimensions, resources

manifest at different lifespan stages, as indicated in Table 38.

Table 38 Examples of resources across a lifespan perspective Resource Childhood Early

education Late education

Career CoP*

Hardware (for example, computer)

Software x x x x Magazines and print media x x Chat programs x x Websites (industry) x Financial (monetary) x

*examples of resources used by Communities of Practice (CoP)

Resources that provided information relevant to employment opportunities

appeared pertinent in late education and early career. Participants utilised the internet for

job hunting, including online industry forums, online recruitment and a Queensland

government website (M7)68F

72. Several women and a male freelancer noted the value of

gaining information from these resources. Secondary data sources also support the

contention that accessing social networking sites such as LinkedIn provides access to the

“occasional job postings” (Appendix 17, Item 9):

There are online forums as well. There’s a lot of those where you can get the latest

information of what’s happening industry wide. You can have people who might be in

the industry anonymously having a bitch about what’s going on within a certain

company. You can get inside information that way which can be helpful to give you a

bit of an insight. (S1)

Accurate information regarding working in the industry was required when

participants were making career decisions while still at school. However, the male

72 Seek.com is an online recruitment websites. The government website is: http://jobs.qld.gov.au/.

freelancer noted that, in general, students would “definitely not know about such

resources ... I don’t want to stereotype, but they’re young, they don’t really have an idea”

(S1). Secondary data supports the finding that there has been a lack of information

regarding career choices for the games industry. For example, a blogger recounted the

comments that two women employed in the Brisbane games industry made at a games

conference event:

Regarding the lack of women with careers in gaming, it became clear that the industry is not exactly well publicised as a career choice - both Penny and Hannah said they both had no idea you could even HAVE a job like that, until they basically fell into it. (Appendix 17, Item 43) Financial resources were indirectly identified by two participants. One participant

(M4) highlighted that her mother paid for her study and bought her a computer; in her

mid 20s she was waitressing and could not afford these resources. The mother’s financial

assistance was seen as instrumental to the participant’s ability to pursue a qualification in

multimedia. Another noted how difficult it had been to pay for her education. Thus, it

can be seen that access to financial resources can foster access to other types of

resources.

5.2.2 THE PERSON(P) Having described the Environment

(E) category or dimension, the focus of the

findings now shifts to the second category in Framework 1, that of the Person (P).

Findings indicate a person can be described by their social identity. As Figure 14

illustrates, the dimension of social identity is comprised of two key properties: 1)

gendered identity, and 2) the DCI occupational identity. Age and ethnicity were noted to

a lesser extent.

Figure 12: Refined Person (P) category

1) GENDER PROPERTY

Findings indicate that the participants’ socially recognisable gender could be an

influence. The female participants often felt judged against gender stereotypes relating to

women’s capacity and interest in technology and DCI careers. For example, participants

Dimension of social identity

Properties

reflected that educational practices socialised women away from technology-related

careers in the education context:

It's becoming what woodwork and metalwork became and home EC [economics] for

girls and woodwork and metalwork for boys and now it's sort of like film studies and

multimedia for boys and creative writing and art for girls and it's still that divide. (G1)

Secondary sources of data indicate that even strategies that aim to foster

women’s involvement may at times use language that reinforces stereotypes. For

example, an email announcing a technology scholarship for girls refers to the girls as

“young ladies” in the subject line (Appendix 17, Item 45). Even ‘research’ may carry a

message that implies that ‘normal’ women are not suitable to technology-related careers.

For example, an article reports that “women in technical careers have ‘male’ brains”

(Appendix 17, Item 46). An Australian professional gamer noted in an online article that

women face challenges in participating as gamers because they are identified as being

‘women’: “Gender sort of is a big deal and a lot of girls take flak not for anything except

being girls” (Appendix 17, Item 22). A blogger reporting on a public games forum held

in Brisbane (2009) stated that she is pleased that one of the presenters made the point that

she is a “normal” woman who also enjoys playing games.

Thus, a tension appears to manifest between what was described as ‘normal’

women (linked to gender stereotypes) and those working in the DCI. Participants

distanced themselves from certain types of women. This distancing was interpreted by

the researcher from women’s comments that they were not like other women: “You're not

quite the same as the other girls” and “you’re not really into the makeup and things” (M2).

‘Other women’ were also defined by one participant as not having technical skills: “I

can’t imagine a really girly, girl working here but if she had the right skills I think they'd

[male colleagues] just love it” (G4). Even the “power-suited” women in the IT industry

were seen as “a little weird” (M3). The sense of distance one participant felt from the

image of these women in IT influenced her pursuing a career in games production. It was

“one of the reasons why I got into games” (M8).

This distance from ‘normal’ women may explain in part why several participants

did not personally identify with either masculine or feminine traits, but rather, as one

participant suggested, she was “stuck in between” (M2). Others made a point that gender

was not an important aspect in regards to working in the industry: “In the games industry

it doesn’t matter if you’re a boy or a girl; it matters how good you are” (G7). One

participant stated: “I don’t even really notice that I'm a different gender, it's a non issue”

(G4). Another noted that when she first started her university education in multimedia,

gender was not a concern: “I never came to a degree thinking it’s going be male or

female. I never came with that perception in my mind” (M7). Thus, the evidence suggests

that the social stereotypes women face influence their participation. However, the women

in the case study had different, individual experiences of, and responses to, gender

stereotyping.

Most participants indicated that the stereotyping was generally manageable. This

is exemplified by the following example where the only female games programmer says

that although people seem “shocked” or “surprised” to find out she is a “female

programmer” (particularly when she attends interviews), she downplays its significance

and describes it as “just, small, no big deal” (G3). For others, however, their gendered

identity became a constraining influence when it surfaced in their day-to-day interactions

and ‘triggered’ participants’ feeling as though they were being judged against a

stereotypical gender characteristic; for example: “We have lots of fun just sitting in the

office and having a good time but there's just some occasions where I do feel like...that's

right I'm a girl” (G1). When women experienced certain conditions that reminded them of

the stereotypes surrounding their gender, there were a number of consequences. These

included realising that they were in a “minority”, that they felt “different” to their male

colleagues, and that they felt judged (or they even judged themselves) according to the

stereotypes surrounding them. One participant recounted her experience in Japan (as she

was required to travel there for her role). She suggested that such constraints could be

worse because:

Over there, there's no one female as a technologist…So sometimes when I go over

there representing Europe, I feel that they are uncomfortable dealing with

me...sometimes it's a bit grating. I found often when going to meetings with one of my

subordinates, who is a guy, and meet people and they just assume that I'm like the tea

lady or something and not even look me in the eye. (G8)

The stereotypes that suggest that women do not have technical proficiency often

left women facing extra pressure when interacting with colleagues:

When I deal with technical guys they almost want to test you and they assume you

have no idea about anything technical…They don't just take you at face value,

whereas they would a guy. They assume a guy knows. I think it's very difficult to be a

female in this industry because potentially you have to work so much harder to get

noticed and to get the same level. (G8)

This placed women in a position where they felt that they needed to prove their

capability as DCI professionals: “We're there so we might as well show that we can do it

just as well as they can…It's not necessary but I think just when you can prove

everything they can do…it's kind of just gratifying” (G5). For others, there were some

advantages of not being perceived as not being technically proficient because of their

gender:

I never get asked to help out with all the crap jobs that the guys get stuck doing [...]I'm

not expected to learn all the technical non-web dev stuff like working with servers[…]

the (male) programmer (with no prior knowledge) is constantly hassled to help out. On

one hand it kind of bugs me that I'm instantly excluded just because I'm a girl, but it is

good because then I'm not constantly being interrupted. (M6)

In addition to women pulling away from a stereotypical gendered identity, many

experienced the pressure to fit in with the male majority–often referred to as the ‘boys

club’: “As I am the only female on my team, it can at times feel like a ‘boys club’” (G1);

and “it is gender centric, in that it’s very much a boys club" (G7). Men are seen as being

capable with technology and thus they become the norm or measure in the work

environment. If a woman wants to “make a game”, she just has to be “as good as the

boys next to them, that’s it” (G7). One participant suggested that there were two types of

women in the games industry. Those women that fitted into the boys club were the “hard

nuts” who “all sort of get in there and play poker and have the odd swearing match and

it’s fun” (G7). Being one of the boys means “you’re part of the group” (G7). Less was said

about the other women’s approach; however, it was suggested it involved asking for

“help all the time” (G7). Regardless of the approach, the participant suggested that

women responded by adapting to the characteristics of the environment: “So I think it

does attract a certain type of female sort of personality type but I think you probably find

that a lot of women actually just sort of deal with things that they don’t like probably a

little bit better than men, like you evolve to their way of thinking” (G7). Similarly, another

participant suggested that women adapt to the conditions they find themselves in: “It can

be difficult sometimes working in a very male dominated industry. Sometimes you just

get used to it. Your whole career you get used to it” (G8).

Participants recognised inaccuracies in stereotypes. Stereotypes could be both

“valid” and not offer “an accurate picture” (G10); and “Sometimes there's some truth in

stereotypes but there's also a lot of grey area” (G4). Participants’ accounts suggested

gender stereotypes were more pervasive than occupation stereotypes. One participant

noted she held stereotypes about certain occupational roles; for example, that male

programmers were ‘geeky’. Reflexively, she articulated her awareness of the

inaccuracies of the stereotypes; not all male programmers were unattractive: “There's a

couple that are but then there’s a lot that aren't”. “There's this idea that all male

programmers are going to be overweight and big glasses” when in reality they are “good

looking” and “a lot are body builders and surfers” (G4). The participant perceived that not

all programmers were geeky, but that most programmers were male. Further

investigation of the influence of gender stereotypes is warranted.

2) OCCUPATION PROPERTY

Participants identified three characteristics of the DCI occupational identity.

There is: 1) the passionate worker, 2) the technical and creative worker, and 3) the geek.

1) Both the female DCI professionals and stakeholders noted that DCI

professionals are passionate. One suggested that the only reason you would enter the

industry is “because you've got a passion for games” (G1). Another suggested that “most

people end up in multimedia because they have a passion for it” (M4). Participants

valued working with “like-minded individuals who are passionate” (G6). Participants also

suggested that being passionate was as important to employers as the skills an individual

possessed. Secondary sources of data further support this view; for example, several job

advertisements for roles in the Australian DCI indicate that suitable candidates would be

“passionate” and work alongside other “passionate people” (See Appendix 17, Items

21and 24). In the international IGDA diversity survey, several respondents mentioned the

need for passionate workers. One claimed that: “Overall, positions in the industry are

typically filled with passionate individuals that have competence in their field”

(Appendix 17, Item 28-F, 39, White, Uni, USA). Another believed that “Game

development shouldn't be a job but a passion. I don't have any problem with girls

working as game developers, as long as they have a passion for it” (Appendix 17, Item

28-M, 26, White, Ma, Belgium). In general, both women and men could be passionate: “I

love games and creating games and I don't see any barriers because I am a girl. I think if

you are passionate something, then you go after it” (G6). However, two stakeholders

suggested women may be perceived as not being as passionate as men. Passion was often

linked to spending long hours developing skills. It was viewed that if women had not had

as much exposure to resources such as games in their youth, then they may not be as

passionate as a male. Although passion was an important characteristic of the DCI

worker, women faced stereotypes regarding their capability to be passionate.

2) The second characteristic of the DCI professional is that they possess both

technical and creative skills. However, women faced challenges in being perceived as

having technical skills due to the gender stereotypes. Women were not associated with

technical roles such as programming: “I think there's more stereotypes against female

programmers. I've only met one other female programmer…but quite a few artists and I

think the artists face less stereotypes and just more the environment issues like being in

a boys club and that sort of thing” (G3). These stereotypes, may explain why there are

fewer women in roles such as programming that are considered highly technical.

Whereas programming roles were associated with male workers, females were perceived

as being better at administrative tasks and project management; this lead to a work

environment where “men did programming and women paperwork” (M2). These

perceptions of the genders were perpetuated by the “stereotypes and jokes floating

around” suggesting such things as women being good at the “communication side of

anything”, and men being “terrible” at communicating (M2). A stakeholder described how

he wrote different recruitment advertisements to ensure equitable numbers of men and

women applied for his interactive development team. He believed women would apply

for project management roles and men the technical roles, such as Flash development.73

He suggested that women might see communication-based roles as a career option, as

communication skills are stereotypically linked to women. This may explain why

multimedia organisations were perceived as presenting fewer barriers to women than

games organisations: multimedia had “evolved more towards communication media and

attracts a broader spectrum of people”, whereas “games” and “animation” is more a

“production environment not a communication environment” (S3). Participants also

identified that even where a women’s primary role was technical, they often undertook

administrative tasks: “Maybe this is where the ideal of the sexes come in but women take

on a lot more of the administrative sort of roles” (G7). One participant in what might be

defined as a creative role (as an artist), even noted that she took on more administrative

tasks because:

I’m a girl...because it’s something I’ve done before but also something that I seem to

be a bit better at than the chap who sits next to me. So we seem to – it’s not surprising

that we have what we do as an artist but then we’ll take on some administrative tasks

whereas a male might not take on the administrative areas. (G7)

Thus, findings indicate that women face occupational segregation in the DCI.

This segregation could be amplified because the career pathways in the DCI

encourage distinctions between technical and non-technical skills. One participant

lamented the fact that she had to choose either a technical or managerial role, as “there’s

no in between” (M7). Although technical roles appeared valued, one stakeholder (an

employer) pointed out that entry level programmers of packages such as Flash69F

74 were

seen as dispensable. He went on to say that he would actively discourage his own son

from pursuing a Flash developer career path as it had limited career progression options,

a capped salary, and because entry level roles did not offer any level of control in the

production process. This stakeholder suggested that women were at an advantage

because they pursued non-technical career pathways.

73 Adobe Flash is software that requires knowledge of the Actionscript programming language. 74 In the late 2000s, Flash was a key software package for multimedia product development.

Findings also reveal that the stereotypes for DCI programmers and ICT

programmers differ. One participant described how multimedia programmers are “very,

very different than those working in IT”. She described those in IT as being “just a

different breed…I think the way their brain works or something is different than it does

with multimedia, not just with my experience and my exposure but in teaching them and

with working with them” (M7). She made a particular point that IT programmers may not

function well in a team environment because their “communication skills are zero”.

Communication skills were highly valued within the DCI, particularly as labour is often

organised in teams. Thus, findings indicate that although technical skills are often cited

as a trait of DCI workers, the nature of technical roles varies and other skills are valued

in the DCI.

3) The third characteristic of the DCI worker is that they are strongly associated

with the ‘geek’ identity. Being a geek in general was not seen as appealing. Geeks were

described as “slightly unpopular, no life, no girlfriends” with “dubious body hair” (S1).

There is a “perception of strange geeks sitting all night playing games and absolutely

unable to be social in anyway”, which leads to the industry being a “place that's kinda of

uninviting” (M1).

The reason Workforce diversity isn't happening is due to the negative stereotypical image of gamers. If gaming was 'cooler' and games appealed to a wider audience then people from that market would be interested in joining the industry. (Appendix 17, Item 28- M, 19, White, HS, Australia).

Industry and education initiatives have tried to change these negatives

connotations. One university program had tried to “sex up the image of the program now

in schools” as the DCI had no “sexy geeks” (S2). Similarly, secondary data indicates that

there are initiatives that try to redefine the image of the geek. The Sexiest Geek Alive 70

pageant promotes the “sexiness of geekiness” (both male and female), thus recognising

that IT is “not one of the sexier professions”71F

76. Other initiatives promote ‘geek chic’

(Appendix 17, Item 44), a term suggesting that being a geek is fashionable. Being a geek

did have appealing aspects when associated with technical capability. This may explain

why one participant described herself and her colleagues as “proud geeks” (M7). The

geek is associated with technical ability in most programming skills: “They see them as

geeks in terms of they think it’s programming” (M7). Like passion and technical ability,

the geek is also associated with the male gender. Participants referred to male colleagues

as ‘geeks’. However, participants also recognised that they themselves viewed their

colleagues stereotypically: “So I don’t know to be honest I probably see them through

75Sexiest Geek Alive pageant at http://archive.salon.com/tech/log/2000/03/07/sexiest_geek/index.html 76 As reported online in popular media at http://news.ninemsn.com.au/article.aspx?id=284665

some stereotypes” (M1). The influence of stereotypes surrounding identities such as the

geek warrants further investigation.

3) AGE AND ETHNICITY PROPERTIES

Along with gender and occupation characteristics, age and ethnicity was noted,

but to a much lesser extent.

Findings indicate that participants in the DCI are young, or at least it is perceived

as a young person’s industry: “We are a very young group; so I’m 23 and like I know that

a lot of us are in our 20s” (M8). Similarly, the IDGA Diversity Survey suggests that the

DCI workforce is comprised of people under the age of forty: “Very few female game

developers…Age diversity is pretty good, but there are few over 40 that I've seen”

(Appendix 17, Item 28, M, 37, White, HS, USA).

One participant, who was aged thirty, identified that both her gender and her age

made her feel different to the majority of workers in her organisation. She noted that

although she was in her early career (having recently left her teaching career), she was

as old as her manager and older than a majority of her peers. “On a superficial

level…you’re at a disadvantage if you decide later on that it's an industry that you’d like

to be a part of” (G2). Although perceived as comprising a young workforce, the

industry faces challenges as this workforce ages: “When I look at when [partner’s

name] started and he’s been in the industry for 13 or 14 years now, most of the people

were between the ages of 19 and 25. And now there are people having their 40th

birthdays. So it’s grown up. It’s definitely grown up” (G7). Secondary data suggests that

the ICT industry faces similar issues surrounding ageism in its workforce (ACS, 2010).

These initial findings suggest that the properties of gender, occupation and age (within

the social dimension) may influence women’s participation and warrant further

consideration.

The ethnicity of women in programming was mentioned by two participants. One

participant (of Asian heritage) offered an explanation of why there were a higher number

of Asian students in many of the Australian university courses related to IT. She said that

her friends knew that by choosing to study an IT course, they would have a better chance

of emigrating to Australia. She recalled how she had been asked to help tutor a new

arrival because “he picked a Master degree in IT [Information Technology] and he doesn't

even know what IT is about … it's just because of the migration policy”. She continued by

saying that, for female migrants, “IT will be more attractive no matter if they are

interested or not” (M2). Thus, the findings indicate that influences such as government

migration policy and ethnicity manifest in complex combinations.

5.2.3 THE INTERACTION (I) The literature review identified the need to investigate the interactions between

the two entities of Environment (E) and Person (P). The previous analysis of empirical

data resulted in a rich description of the first two categories (E and P) in Framework 1.

Although this process sensitises the researcher to interactions between those categories,

further conceptualisation of the interaction (I) category using inductive analysis reached

an analytical impasse. To move the research process beyond this impasse, further

analysis of the empirical data focused on identifying specific events of interaction.

Although Danermark et al. (2002) did not explicitly focus on the analysis of events, this

analysis can be seen as aligning to Stage 2 where entities are further ‘resolved’. The

choice to focus on events is supported by the Critical Realist ontology underpinning the

research. Bhaskar suggested that there is a need to identify the events in the ‘actual’

domain of reality (as discussed in Section 3.3.2).

EVENTS OF INTERACTION

The 18 women interviewed recounted

several events that influenced their

participation. An event is defined as ‘a memorable or significant occurrence for the

participant at a particular temporal or spatial point’. In exploring the E and P categories,

the findings do, to some extent, describe several events. However, the emphasis has

been on a description of the characteristics of the environment and person, and not on

events. In addition, many of the events–such as the dot.com crash of 2000 (tech-bubble),

and the global financial crisis (GFC) of 2009–have manifested at the macro rather than

micro (organisational level). This section reveals the events at the micro level. Rather

than natural or historical events, the emphasis is on the analysis of the everyday events

women experience in the DCI organisation. Nevertheless, there is some reference to

events that occur at different points of the lifespan perspective, at the macro and other

micro contexts (such as school); this is because the overall aim of the study is for holistic

understanding.

These events manifest at both an individual and group level.For example, at the

group level, each of the participants highlighted that the ‘event’ of entering the industry

was an integral influence on participation. At the individual level, the actual experience

of entering the industry differed for each person. Group level inference lead to the

conceptualisation of 10 events that manifest in the interaction between an individual and

their environment. These 10 events are presented in Table 39 in a way that reflects when

they are most likely to occur across a typical lifespan. This does not imply that these

events must occur in this sequence. For example, as the first event ‘Learning recipes’

identifies, participants may learn skills in their youth and also face the need to maintain

those skill in the adult careers. Empirical evidence supporting this conceptualisation of

the 10 events of interaction is now presented.

Table 39 The 10 Events of Interaction manifesting in the ‘actual’ domain of reality Childhood/ education Career Future career

1) Learning Recipes 4) Ladies Lunches 9) Motherhood

2) Getting In 3) Just Like TV

5) Muffins and Men 6) Catch 22

10) Greener Pastures

7) Mind Your Manners

8) Are You the Entertainment?

1) LEARNING RECIPES

The first event, ‘Learning Recipes’ recognises events that foster skill

development opportunities. Several participants indicated that learning computer-related

skills in their childhood fostered their later participation in DCI education and work

contexts. One participant recalled that, as a young girl, she had gained skills in

programming by following the coding ‘recipes’ in her father’s computer magazines (G2),

which then helped her develop a sense of being skilled in using software. She highlighted

that it was not necessarily any direct intervention from her father, but rather access to a

resource that helped her learn coding: “My dad would get commodore 64 magazines and

so I'd just copy out programs and write games cause in the same way that recipe books

give you recipe” (G2). The participant also noted the link between such childhood

experiences to the skills her fellow ‘male’ colleagues had acquired; these males held all

the senior programming roles in her organisation:

There a couple of work environments that are just males and I don’t know if they'll ever

get a female [which environments are those] The engine for example, just umm very

senior programmers that have been programming for fun since they were a kid. (G2)

Beyond childhood experiences, participants also noted the importance of

developing and maintaining skills in the education and workplace contexts. One

participant, who worked in a public service/government organisation, felt a negative of

their work environment lay in the limitations surrounding skill maintenance or

development: “For me not to develop anything for half a year is really, ready bad skill-

wise. ... [...]So yeah, I think that’s my biggest hurdle ... I have no way in government to

develop my skills [...] but then work-life balance” (M7). Developing technical skills

enabled women to challenge those stereotypes that suggest that females do not possess

technical capability. When women mastered the technical skills associated with the DCI

occupational roles, they developed a sense of confidence: “I'd never really done any front

end interactive scripting before I came here so I had to pick up really quickly and I was

just like, oh, it's actually not as hard as I thought” (M5). Maintaining technical skills often

demanded a great deal of commitment from participants, and this was seen as a negative

influence on participation.

2) GETTING IN

The second event, ‘Getting In’, recognises events that foster participants’ entry

into the industry. Several participants noted that one of the most difficult aspects of

participating was in gaining initial access to the career pathway: “I think it’s very hard to

get that first foot in for anybody, especially women” (G5). Secondary sources, such as

interviews of women working in web design in the UK also indicate that access into the

industry presents a particular challenge to women (Appendix 17, Item 52). Table 40

summarises the variety of ways participants gained entry into the DCI. It is noteworthy to

recognise that a majority of the participants in the study had a personal contact facilitate

their entry into the industry, in contrast to only a few that had replied to traditional or

online recruitment advertisements.

Table 40 Female interactive content creators’ access to the DCI Method of entry into the industry/current role G1 Approached a mentor who offered work experience, then worked with boyfriend in

same organisation

G2 Met DCI workers while working in a café; formed a good relationship before applying for a position at the organisation

G3 Applied for first position; access to second may have been assisted by knowing people at the organisation; moved to second organisation with partner

G4 Boyfriend offered a job at the organisation; facilitated participant’s entry when opportunity arose

G6 Responded to online job advertisement

G8 Freelanced as a web designer M1 Was freelancing and current employer was a client

M10 Studied marketing and got an initial position in an advertising agency; then moved into DCI role

M3 Helped by student program which facilitated access into studio environment; informed of available position by fellow female student (who had gained entry)

M4 Helped by student program which facilitated access into studio environment; networked with clients to change occupation; current partner in multimedia

M5 Male friend put name forward at the organisation; previously had numerous job applications rejections

M6 Informed of opportunity by flat mate; current partner in games organisation M7 Informed of position by friend M8 Informed of position by university lecturer

Findings indicate that a person’s skills, serendipitous opportunities, industry

recruitment practices, and access to networks could all influence an individual’s entry

into the DCI. Access into the industry was seen by participants as being highly

competitive, particularly for entry level roles; however, access into certain roles was seen

as easier than for others. For example, QA (Quality Assurance) was seen as a “really

quite easy entry level position to get into” (G6). In contrast, other roles, such as games

design, were “pretty rare” and “hard to find” (G6). Previous experience and skills were

highly valued by recruiters. Certain roles such as “artists and animators and even level

designers” (G2) were recruited on the strength of a “show real or the portfolio” (G2).

However, this experience did not have to be with a DCI organisation. Experience could

arise from tinkering at home and building websites.

For other participants, gaining access into the DCI involved serendipity. As one

participant reflected, “I didn’t mean to get into games” (G7). Ambiguity surrounding roles

and career paths in the DCI appear to lead to the situation where “people kinda fall over

their careers” (S3). For example, one participant recounted that while working in a coffee

shop, she befriended the male workers from a local games company who came in for

coffee: “I became quite good friends with them” (G2). She had formed a strong social

relationship, even having dinners with them, before she considered a job at the

organisation. It was her exposure to these workers that led to her interest in the industry.

She noted that while serving them in the coffee shop, she would “hear a lot about what

they were up to… how work was going, what kind of hours they were working, all the

social kind of things…they always seemed to really enjoy what they were doing” (G2).

Another participant, an AI programmer, described her ‘luck’ in securing her first job

when the organisation took the bold step to “cast a wider net than usual” (G3) and hired

her as a junior, when they “very rarely hire junior positions”. In addition, she noted: “I was

lucky that the company I applied at was umm very blind to my gender because I know

they’re not all like that” (G3). Had there been no opportunity for exposure to the industry

or entry level positions, these participants felt they would not have entered a career in the

Recruitment practices influenced participants’ access into the industry. Few of

the women had applied for their positions through direct application. Indeed, some

suggested that they had faced many rejections following that process. One participant

who spent a period of time assisting the HR person in her organisation noted the

discrepancy: “All the time heaps and heaps of emails and cv's in the mail for games

testers, one female applies and there are hundreds and hundreds of guys everyday…

the whole time I was there, one female” (G5). Another female participant described the

difficulty she had in recruiting females even when she “purposely looked for females... I

didn't get any applicants a lot of the time...Or if I did they were too junior...when I did get

them... it's probably one in 10 is female” (G8). Secondary sources of data support the fact

that there are few women applying for specific roles such as programming: “We need

better ways of involving more women[...] this is difficult when 99 of 100 programmer

resumes received are from men” (Appendix 17, Item 28, M, 31, White, disabled, Uni,

USA). However, this contrasts with the preferred recruitment channels identified by

games organisations internationally: Direct application (100%), industry websites (78%),

networking events (56%) and use of recruitment firms (44%) (Appendix 17, Item 54).

One participant who worked in the public sector viewed the written application process

as leading to an “equal” number of “male and female” staff (M7). Similarly, another

participant suggested women performed better in responding to the detailed written

selection criteria required of applicants and men were better at “talking or selling

themselves” in an interview (M7).

The manner in which job advertisements are presented may influence women’s

entry into the industry. One employer identified that his staff had reasonably equitable

numbers between genders: “For the last ten years I've always had out of a crew of

between 5 to 15, 1/3 to 1/2 women- in technical roles (S3). However, the interview

revealed that particular recruitment practices were being used by the employer to recruit

females. The employer described that if he advertised with an emphasis on production or

management skills, he could be assured he would have a majority of female applicants. If

he advertised for technical skills, it would be a majority of males applying. He believed

women were making the decision to pursue non-technical roles before applying:

They're making the decision before they come here…that there are better career

opportunities and more ummm control over their lives...or whatever I'm not really sure

what drives the decision but to apply for project related roles as opposed to

programming roles. (S3)

Access to networks could foster entry into the industry. One participant noted, in

reference to securing her position: “I didn’t even really look. Just [friend’s name] is my

friend at [organisation] and I guess [organisation] told her they were looking for someone

and then my friend told me” (M7). Indeed, for several participants and stakeholders,

networks appeared to be the preferred method of recruitment: “I think I found that later in

my career that I got all my work and jobs through word of mouth, but in the beginning you

don’t have that so I didn't get offered any jobs.... it was all word of mouth, never

advertised” (G8). However, as a male freelancer notes, these networks were often

comprised of males and, thus, career opportunities involve networking with males: “Jobs

for mates and that sort of thing, so it's a bit of a boys club” (S1).

3) JUST L IKE TV

The third event, ‘Just Like TV’, recognises events that perpetuate inaccurate

perceptions of both the DCI and women. Participants recalled how inaccurate

information perpetuated by the media (television, newspapers, and websites) resulted in

women not being interested in DCI occupational roles:“And so you have to really scrape

all the crap off and sort of see and find out who these people are and it would be a good

boost, a good PR exercise for the industry” (M5). People need to sort of dig a little bit

deeper, like go beyond the press release before they start believing how viable a

particular career path is (S1). Participants emphasised the need for an accurate portrayal

of women working in the industry. It's about representing the industry in a way

people are going to relate to it (M9).

Participants identified the specific need for industry or education professionals to

talk to school-aged children to inform them of career choices: “that’s when you go into

schools and actually talk to them about what it is” (M7). This information was needed to

assess a fit between roles and personality: “I think it’s a matter of letting the girl know that

these are all the different jobs you can have, which one suits your personality, which

one's going to be the most rewarding as a career” (G5). Many stereotypes perpetuated

by the media, such as Maths being boring, were inaccurate:

When you’re at school you don’t know…you thought that the kind of mind that you

needed or the kind of brain was more like a maths-y brain or something and you’re like,

I couldn’t possibly, it’s not me, it’s a bit boring. But if you took it one step further and

looked at the course details and chatted to someone, you’d probably get a different

spin on what these jobs are like. (M9)

Stakeholders also emphasised the need for accurate industry information to be

provided at a time when girls may be formulating career aspirations:

The one thing I’d like to see at happen at the university level or high school level is to

get the message across that the interactive industry is a part of the communication

industry and it's an interesting variation…It kinda needs to happen somewhere before

the education system, somewhere between high school and sort of the higher

education system. (S3)

4) LADIES LUNCHES

The fourth event, ‘Ladies’ Lunches’, recognises events that can make an

individual feel isolated as a result of being different to their colleagues and peers.

Participants described feeling different, lonely, and isolated as a consequence of being in

the gendered minority in workplace and education settings. Women in one games

production organisation (Organisation B, see Table 23) responded by organising a social

event that facilitated their professional collaboration and fostered personal connection

with other workers. The event, ‘Ladies’ Lunches’, was a regular lunchtime meeting, to

which female staff and even colleagues’ wives were invited. In addition, several women

had actively pursued social events as a way to combat the sense of isolation that

developed from their being in the gendered minority. These social events were also seen

as an integral way to develop contacts and skills: “At my first job I always went to lunch

with the guys for the first year at least because I knew they talked about work at lunch

and I knew I needed that experience” (G3). Other ways of coping with a sense of

isolation included signing up to online communities.

The female programmer (who worked in a multimedia organisation) reflected

that she also felt lonely when she was a university student, as there were few women in

her IT degree. Secondary sources also support the finding that women are a minority in

some university IT programs, with an online website capturing this comment: “Put some

compulsory IT classes in all degrees for chicks, it’s getting a bit lonely in the IT lectures”

(See Appendix 17; Item 42). However, she noted her loneliness was minimised during

her undergraduate degree because the IT classes included those female students from

other degrees that were picking up IT subjects. It was only when she was in her Masters

of IT course (and the only female in the class) that she felt as though she was in the

minority. She recollected that, on the first day, the male students made a joke about her

arriving late; they had thought there would be no women in their class. However, she did

not feel that the learning environment was difficult–primarily because she was only

expected to interact with the students for a couple of hours at a time; however, “if that

was going to happen every day then it might be a different issue” (M3).

Later in life, this same participant found herself in a work environment where she

was the only female of a team of about 10 programmers. However, she noted she did not

feel isolated as she could mix freely with other females (mostly from the web design and

coding team) over tea and lunch breaks. She noted that even though she didn’t think it

was a problem that she was the only female programmer, she suggested that was because

of the nature of the company she worked with as it was a “very close together company”.

The size (number of employees) of the organisation or department was important: “If it's

in a bigger company I think it may be a problem because every department just isolated

and I will find out I am the only female sitting in a big room” (M3).

5) MUFFINS AND MEN

The fifth event, ‘Muffins and Men’, recognises the events where gender

stereotypes surface in workplace practices to influence participation. Earlier findings

identify that women face stereotypes that create a sense of tension between their

occupational and gendered identity.

One participant recounted a specific event where she interacted with her

colleagues and managers within a games production organisation. The event had

happened only the day before the interview and the participant appeared visibly upset.

She described how it appeared that, because she was a woman, male colleagues in the

QA department had assumed she would be responsible for the paperwork involved in a

particular project.72F

77 She suggested their attitude was “can you deal with this, it's like yuck

that's like girls stuff” (G1). The participant recognised that the task was not her

responsibility; however, in seeing that the task would not be completed if she didn’t do it,

she completed it with the help of another person (a male). In return for his assistance she

gave him with a box of muffins. A complaint was made against her for not completing

the paperwork on her own. When management became involved, she felt they supported

her, but for the “wrong reasons” (G1). Rather than clarifying that the paperwork task was

the responsibility of the QA department, management implied that, if it happened again,

the participant could simply supply more muffins.

6) THE CATCH 22

The sixth event, ‘The Catch 22’ recognises events where gender stereotypes

influence a women’s capacity to negotiate within the workplace. Findings indicate that

women may face difficulties in negotiating salaries or employment contracts due to

gender-related stereotypes.

One participant described a work environment where she felt she knew as much

as a male colleague in a comparable position, who was earning a “lot more money” than

she was. When asked why she thought this was so, she replied: “It's just one of those

things, you think, I know at the end of the day you think 'how the hell', the worlds

obviously you know, screwed” (M2). The participant attributed pay inequity to the

difficulties women faced when negotiating salaries and employment contracts. She was

aware that the attitudes towards women often meant women did not negotiate as

forcefully as men: “I do firmly believe that women are more sheepish when it comes to

salary and promotion and male bosses will take advantage of this if you let them” (M2).

She described how it was not unusual for males to approach salary negotiations with

their employer with ‘cockiness and bravado’; however, “in society it’s not accepted for

women to be that way” (M2). She identified that even she would feel uncomfortable if a

woman acted in a similar ‘cocky’ fashion to a male. Her response in that situation would

be “what the!” However, discursively, she noted she was in a “catch 22” situation: she

wanted an equitable salary but felt constrained when it came to negotiating it because of

social norms and stereotypes surrounding the female identity. She recalled how she was

“totally freaking out” when she heard of AWAs being proposed, mostly because the

AWAs leave workplace negotiation in the hands of the individual, in comparison to the

77 As an aside, the participant has since left the company, although it is not evident why.

provisions of workplace agreements or union-based awards.7:

78 “I could of died when the

gov brought in AWA's - cause most females are not good at negotiating” (M2).

Secondary sources, such as the Queensland Pay Equity: Time to Act (2007)

inquiry recognise the influence of inequitable remuneration and suggest that females on

awards earn more than males on awards (Submission 58). Similarly, ABS surveys also

indicate that women can be paid lower wages in the ICT Industry (See Appendix 17 Item

36). 65F

79 As of July 2012, the ABS surveys do not provide pay equity data by gender for the

DCI. Compounding the difficulties in negotiating salaries is the ambiguity surrounding

them (as previously noted). Difficulties were amplified in organisations where there was

a culture of secrecy surrounding remuneration:

I didn't like was the sense of secrecy about the wages […] it was just very secret hush,

hush business and I really felt that deteriorate the moral and then once that goes it's

like you all become whinging bitches basically.(M2)

The need for self-confidence was an important aspect of negotiations. Another

female acknowledged that she may have secured a management position with a

prestigious games development organisation because “I probably blagged it in the

interview” (G8). As an employer, she recognised that “guys definitely talk themselves up a

lot more”:

Certainly, in the technological aspects they will talk themselves up. Girls will be more

truthful. I think you come to really see through all of that bullshit as well - you hope. But

the guys definitely, yeah, promote themselves more. (G8)

Secondary sources of data also indicate that women need to promote their value

to their organisation. An email from an IT industry group states women should “be

prepared to explain [their] value” (Appendix 17, Item 12). Similarly, an email from an

industry organisation Women in IT in the UK (Appendix 17, Item 17) promotes a training

event to teach women how to negotiate and be ‘confident’. Another email from a

LinkedIn group, states “it's necessary for women to negotiate successfully now more

than ever” (Appendix 17, Item 35).

7) MIND YOUR MANNERS

The seventh event, ‘Mind Your Manners’, recognises events where gender

stereotypes influence a women’s professional identity. Findings indicate that when

women face gender stereotypes this can trigger a lowering of ‘confidence’ in their

technical ability.

78 Australian Workplace Agreements (AWAs) were formalised individual agreements negotiated between the employer and employee. They were introduced in 2006; however, new AWAs were banned under the Fair Work Act, 2009. 79 Quarterly Average Weekly Earnings (AWE)79, the annual Employee Earnings, Benefits and Trade Union Membership (EEBTUM),79 and the biennial Employee Earnings and Hours (ETHE)79 surveys; AWE and ETHE are surveys of employers, while EEBTUM is a survey of employees.

Three participants recounted the event of being exposed to swearing. One

participant in the games industry described feeling awkward in her “immediate sphere of

daily life” when her male colleagues seemed to restrain from swearing in front of her,

apologizing if they accidentally did so. She felt frustrated that even after “being together

for so many years” and “feeling comfortable working” the apology came because she was

“one of the only women they see in their day”. Insightfully, the participant suggested her

male colleague’s action was not malicious. However, she described how her colleague’s

actions triggered her own gender stereotypes, making her feel like a ‘girl’. She

wondered: “What else did they think…how else did they modify their behaviour around

her…are they doing other things to accommodate this woman in their presence …and

how far does it go… the way they consider the women in our company” (G2). She

emphasised that this was a minor event, which she only thought of when she was

“ racking my brain to try to think of a negative”.

Another participant identified that the language her male colleagues used, or

refrained from using, could make her feel awkward: “Like if they watch their swearing

around me that just feels weird to me that their treating me differently that they just

assumed I'd be offended” (M6). Another identified that she is “occasionally frustrated by

guys who feel they need to adjust their behaviour” around the women at work, and gave a

specific example of male colleagues restraining from swearing in her presence: “It has

the effect of patronising the females and seems like quite archaic behaviour and leaves

me less comfortable to know that on some level I'm not considered to be a complete

social equal” (G2). Similarly, another noted that such practices could influence

interactions with colleagues and her sense of fitting in:

Where a women who would take offence to a swear word probably would make people

here feel prohibitive how they could interact with each other. Because boys, when

there's a group of boys, tend to be like you 'bloody -----' and I think if they couldn’t

interact with each other in that natural way because there's a women present suddenly

that would introduce gender problems. Whereas most of the women that work here are

kind of tom boys including me so the guys can interact in the way that is culturally done

in a games company. (G4)

The women did not suggest that swearing offended their professional identity.

Rather, the concern lay in the fact that when males refrained from swearing, and treated

the women differently, the gender aspect arose for the women. Participants did not want

to feel different to their peers or make their peers feel uncomfortable: “Although I said

that it annoys me that the humour is quite juvenile at times I also wouldn’t want them to

like change it in front of me, feel uncomfortable around me” (G3). When such practices

reminded the participants they were a ‘girl’, their professional confidence faltered.

Maintaining a confidence in their ability, and peer acknowledgment of their skills, was

seen as an important way to remain feeling professionally valued in the collective. One

participant noted that although her technical ability was intrinsically “gratifying”, she felt

it was important for her to “prove” her technical credibility to her male colleagues. Here,

the emphasis was on being firstly perceived as a valued, skilled member of the team and

then she could add: “as a by-line ‘and I'm a girl!’” (G2). When asked how establishing her

technical proficiency had an influence, she noted that it made the males “more

comfortable in an environment where slowly bit by bit more females are starting to join

the industry” (G2).

8) ARE YOU THE ENTERTAINMENT ?

The eighth event, ‘Are You The Entertainment’, recognises those events that

foster unprofessional male-orientated industry practises. Findings indicate that industry

practices may promote women as entertainment, rather than as skilled professionals. This

was evident in the types of entertainment provided at industry events, which could be

interpreted as being orientated towards males. As one male stakeholder himself observed

at a games conference held in Brisbane, the after dinner entertainment comprised of a

female dance troupe:

Sony...funded that event. But the only girls who were there other than – I can

remember like two or three actual women who were partners of the people involved

there. All the girls who were there were actually paid to be there by Sony and they

came out and did their little performances and stuff. But I think they’re only paid to be

there until 12 or one and as soon as it hit that time there was no females at all. (S1)

Secondary sources support the evidence that women are often employed as the

entertainment at technology-related industry events; for example, ‘Booth Babes’ at the

international E3 games conferences (Appendix 17, Item 7), and the meter maids at the

Microsoft-supported Australian 2010 Tech Ed Conference (Appendix 17, Item 40).

These practices have been questioned by those in the industry and the wider society:

“Need more women in the industry, and in order to do so will need to eliminate ‘boys

only’ practices such as Booth Bimbos” (Appendix 17, Item 28, F, 22, White, HS,

Canada). Backlash led to public apologies by Microsoft who stated that “the use of Meter

Maids was decidedly inappropriate” (See Appendix 17, Item 40). ‘Booth babes’ at the

US-based Electronic Entertainment Expo were banned in 2006 (Adams, 2006). Even

initiatives that aim to foster women’s involvement in IT, such as the ‘Screen Goddess’

calendar (See 6.4.2 for further details), have faced public criticism regarding the

appropriateness of the images, with the Australian Computer Society withdrawing its

support for this initiative (Appendix 17, Item 16). Such male-orientated practices could

alienate women in the work environment. A participant recounted how she was never

invited to the after-work visits to strip clubs when her ‘team’ and US-based publishers

socialised: “They just do things socially that I don’t get invited to” (G3). This was seen as

constraining her career progression, as her male colleagues could establish a “rapport that

I don't have with the publishers” (G3).

9) GREENER PASTURES

The ninth event, ‘Greener Pastures’, recognises events that constrain women’s

career progression. Constraints manifest because of the industry’s general lack of well-

defined career pathways and opportunities for progression, gender stereotypes regarding

women’s occupational strengths, and the value of certain occupational roles.

One stakeholder, an employer (and industry representative) acknowledged that

workers (both men and women) faced difficulties formulating career pathways because

the industry offered unclear career pathways. It was a “challenge to articulate it [career

pathways] even in our own organisation, let alone at an industry level”, and there appears

to be no “clear understanding of tried and tested career path ways” (S3). Both men and

women face limited opportunities for promotion in technical roles. Games, in particular,

could be “an industry where…there seems to be very limited sort of areas for you to

progress in being male or female” (G7). Often, career progression entailed moving into

management roles, rather than furthering one’s technical role:

The longer you stay, the more narrower your career path becomes...you either become

a senior programmer or a technical director. But there’s not usually more than one

technical director in an organisation, whereas there may be 5 or 6 programmers or

otherwise you’ve got to specialise in Iphone development or something like that. (S3)

The stakeholder suggested that career opportunities surrounding certain technical

roles may not be appealing to women:

One of the best Flash developers we've recently had umm is a female and she has,

almost as quickly as she came, graduated beyond that role of as a technical person

towards a project management lead. And I say graduated because to me... it’s actually,

the reason why a lot of guys are Flash developers, technical people […]To be perfectly

honest I'd prefer to be a female going through that role because you actually get more

career opportunity whereas the guys tend to get stuck in fact that they've got to keep

their hands on the tools. Whereas umm the girls I've seen come through are quite

happy to let that go and move onto greener pastures... (S3)

Several participants suggested they would like to have more control over their

careers and life/work balance. For many, this entailed owning their own studio in the

future, because “if you have your own gig you can make the rules” (G3). The employer

attributed the unappealing nature of technical roles, in most part, to the lack of control

over work tasks, the project outcomes, and to life/work balance in general.

Technical expertise to me and I think in the industry is largely a task where people are

told what they need to do ... so to have your women in technical roles will give them

less influence on the outcomes of the project” ...“you get to that point in your life where

you really feel like you need more control over your life therefore you feel like you need

more control over my job. If I'm going to have more control over my job I need to be

become more senior... (S3)

Although women entered the DCI in low level or junior roles (for example, QA

assistant, production coordinator, or technical coordinator), they quickly progressed to

more senior project management roles. An employer noted that “in some respects to be

perfectly honest I see women doing better out of the arrangement as it is” (S3).

Specifically, the employer appeared to place a lower value on certain roles or skills, such

as Flash development. Secondary data suggests women may be pursuing higher paid

roles that are non-technical (Appendix 17, Item 48):

To be perfectly honest, if I've got a guy here at 65 grand as a Flash developer and he's

approaching thirty years old he's at the top of his pay range. He's not going to get

much more he's just going to get older. If I've got women here who are 22, 23 on 65

grand who are at the bottom of their earnings potential they're only going up. And it's

because they’ve chosen...it's where the ceiling lies. The industry pumps out Flash

developers, junior Flash developers all the time. I can keep replacing those guys but

the critical thinking that adds value to the organisation usually comes from girls. (S3)

10) MOTHERHOOD

The tenth event, ‘Motherhood’, recognises that women encounter constraints due

to their parental responsibilities. Findings indicate that the industry does not foster

family-friendly practices.

Although only two participants in the study had children, almost every woman

interviewed mentioned that having a child would present challenges. The event of

becoming a mother was seen as being so significant that several participants suggested

that their careers would end: “I imagine that it would pretty much come to a halt if I did

decide [to have a child]” (G6). Those participants who suggested that they would not

consider stopping work had few strategies for coping in the situation. The AI

programmer indicated that it might be easier to work in another industry. However, she

also noted that the skills she had developed would not transfer easily: “I've specialised a

lot and I've learnt a lot of specific things that I can't apply now” (G3). Participants

indicated that they had concerns for their future participation. For those that had children,

becoming a mother presented the participants with challenge and changes. A games

worker, with a young baby noted: “It’s a bit of a social change. It’s a bit of a financial

change; all sorts of changes” (G7).

Certain occupational roles were seen as being less compatible with becoming a

mother. Management roles were seen by one participant and one employer as offering

women greater flexibility and control over their lives than the technical roles:

when you think about the future […] when I have a family, there’s no way in the world

[you’ll] work in the private sector, no way in the world. You know, that’s unless you’re

maybe in a management position where you have a bit more flexibility. (M7)

However, this contrasts with another participant who had secured a high level, permanent

executive role in a multinational organisation. She stated that she couldn’t “imagine

having a family in that role” (G8) and that it might be easier for women in technical roles.

Thus, both technical and management roles were perceived as presenting women with

challenges.

Several women suggested that they would undertake a part-time work pattern

when they became a mother, in order to achieve ‘life/work balance’: “I think my

profession is actually quite women friendly…I don't have to sacrifice my career, I can

work around it. I can do more freelance art, I can do part-time” (M1). However, for most

of the women, part-time employment did not seem to be an option: “We don’t really have

anyone working part-time. We have casuals and we have fulltime people. We don’t have

part-time people” (M8). Some participants identified that part-time work would not be an

option with their occupational role: “I'd love to do that but try telling a games company

that you’re only going to work four days a week” (G3). The games programmer noted that

in “My specific role I’m not sure, on this game, on this type of role but if I expanded into a

different role and had a really good relationship with the games company” (G3).

Similarly, a production assistant in the games industry did not consider her role as

compatible with part-time employment:

I think being probably an as asset manager and just having that role umm I could do it

part time, I think being a prop builder, or an animator it would be quite easy because

you get your set task and you know that is where it starts and this is where it finishes

this is what it's meant to look like this is what it's meant to do it's sort of black and

white. Whereas what I'm doing it's … you know I need to be around. (G1)

Several women expressed a need for information regarding their options when

they became pregnant. However, it appeared that workplaces offered ambiguous

information. For example, in one work place, two participants had different perceptions

of maternity leave options. One participant said that although the company was “fantastic

with paternity leave but there's no such thing as maternity leave…I'm pretty sure I'll be

one of the first ones to get maternity leave” (G1). This participant appeared to base her

concerns on a previous event she had observed where a women in a similar role, and

(interestingly) who also had a partner working at the same company, had been working

part time until about four months into the pregnancy and “ then just disappeared, she

was just gone”. The other participant believed the company was supportive of new

mothers, even if the industry or actual work conditions were not: “[Organisation name] in

particular is accommodating of new mothers…it’s maybe…long hours, the staring at a

computer screen for hours on end. Maybe that's not the most appealing prospect for a

mother to be” (G2).

The women employed in the public organisations stated their organisations were

more supportive of parenthood, particularly as they offered clear policies and procedures

regarding parental leave: “I think they do really make an effort to be supportive of people

and with families and sort of talking about families” (M8). Across the private

organisations, women felt as though there were few supports for them if they became

parents: “There’s nothing there if I did decide to have a baby” (G1). This often manifested

because of the organisation not having had experience with maternity leave due to the

male majority. Another participant identified that she did not believe her company would

provide “any kind of maternity pay or anything like that. And you know they probably

don't have to because they're all men working in the industry” (G6). The following

comment, from a participant working in a local studio of a large multinational games

company, shows how DCI companies may not be prepared for issues surrounding female

employees becoming mothers:

At the time I fell pregnant there were 90 to 100 staff members and there were six

women and four of them worked in the actual art realm…design. When I fell pregnant I

told my boss and my boss doesn’t have children and he was a little like, oh, I don’t

know how to deal with this. (G7)

The distinct lack of female role models in the industry, particularly those who

were mothers, was seen as an influence. Participants identified the need to talk to

someone who had experience in balancing work and family: “I’d want to talk to other

people who’ve sort of been theret” (M8). Thus, perceptions of incompatibility between

motherhood and career appear not to be based on direct experience (as many did not have

children or know of women who did), but rather on vicarious experience, where the

women had observed the impact of industry demands on their male colleagues who were

trying to manage their family commitments:

In this industry, I would be concerned about having a family just because seeing the

guys at work who are just there all the time and I know they've got babies at home that

they should be spending time with, and they don’t have the opportunity because we're

on deadlines. (G3)

There was a suggestion that colleagues would treat you differently if you had a

child. For example, those with parental responsibilities may be “treated with leniency or

you do what you can” (G2). A participant recounted how, in other industries, she had

observed women trying to downplay their parenting responsibilities. She saw the

weakness of the ‘power-suited’ women as being that they tried to be ‘too strong’:

“Obviously I didn’t have kids then, but when I think back to how they were, these women

in those industries, they didn’t do it. Oh no, no, like we’re proving a point, we are the

woman, we are stronger” (G7). In contrast, she observed that there was little stigma

surrounding men’s family responsibilities: “When these guys go home because their

wife is sick and they’re going to go home and look after the kid, no one cares” (G7).

Secondary sources indicate that employers may be less willing to hire a woman

than a man because of the perception that she will take leave of absence from work. An

interviewee in a leading industry-related magazine replied to the question “Do you feel

that you have ever been held back because of your gender?” in the following manner: “ I

think that many employers are just afraid women will become baby factories and end up

out of the office more than in” (Appendix 17, Item 52). The two women, who had

children, employed a variety of actions to foster their ongoing career participation. Both

emphasised that they actively maintained their skills whilst on maternity leave by

ensuring they had resources such as software at home:

I mean I’d pop in every couple of months and go, shit, look at what they’re doing. I

don’t know if I can do that. So the first thing is I would have to go home and practice. ...

for the first few weeks of my maternity leave, I was actually working from home part

time – maybe 15 hours a week. It wasn’t a huge amount but I was doing it. (G7)

However, there were issues with the practicality of working from home. One

participant described her living situation and how, when she worked from home, her

child kept touching the computer equipment: “It can’t be used for work anymore because

my little boy is constantly just drawn to it and he yanks it and pulls it and it’s a big toy. So

unfortunately it (trying work from home] doesn’t work at all, not when he’s there” (G7).

Both of the participants who had children described the challenges of being a working

mother. However, the DCI was seen as a ‘good industry for women’:

It is a good industry for women. It really is a good industry for women; maybe not the

average woman who wants to wear the power-suit thing and like produce four kids but

be really good at her job and a great multi-tasker, but it lends well to I think having a

family. (G7)

However, the same participant flagged her concerns about being able to continue in her

role if she wanted a second child:

I remember this very well – I was in a meeting and I’d just got back from maternity

leave and my boss said to me it’s so exciting having you back and […] now you’re a

senior a lot more is expected. And I sat there and I thought, yeah, no, I mean I love

challenge, I’ll do it. But the one thing that I actually thought was you know what if I

wanted another child, there’s no way I could stay in this industry, and that’s not to do

with my company in particular, that’s just the industry at large. (G9)

Participants recognised that historical practices surrounding family and work are

changing. For the most part, these changes are a consequence of the male majority

getting older and having family responsibilities: “I mean the industry is definitely growing

up. It’s evolving. So a lot of the guys that I deal with, their wives are having babies” (G7);

and “Like here a lot of the guys are parents so umm I think ---the age of the industry--

they'll be more and more and they'll start figuring out what they’re going to do about

those sorts of things” (G3). However, this aging workforce is potentially replaceable with

the next wave of young workers.

The whole creative industry is definitely getting younger [...] – these kids who are

brilliant... a lot of them don’t have a lot of like major family pressures or they will work

an extra 24 hours a week if they have to because they can. (G7)

5.2.4 REFINEMENT OF FRAMEWORK 1The initial guiding categories offered by Framework 1 of Environment (E),

Person (P), and Interaction (I) between the E and P, have been refined through the

analysis of the empirical data. This refinement led to the development of two models: 1)

the Sphere of Influence (SoI) model, and 2) the Events of Interaction (EoI) model. The

SoI is comprised of the E and P categories, where E is comprised of the four dimensions

of the environment (social, cultural, mediated, and resource-related, as described in

Section 5.2.1), and the P category which is comprised of one dimension–social identity

(as described in Section 5.2.2). The EoI helps describe the ‘I’ category by highlighting 10

events of interaction between the E and P (as described in Section 5.2.3).These two

models, the SoI and the EoI, are contributions emerging from this research. The two

models are discussed in detail in the following chapter.

5.3 FRAMEWORK 2: THEORETICAL (HUMAN AGENCY MTS )

Chapter 3 proposed that a theoretical framework comprised of a Human Agency

Multi-Theory Scaffold (the HAMTS) could help guide data analysis. The HAMTS provides

11 concepts for data analysis (as presented in Table 41). Findings indicate that all of the

concepts were evident in the empirical data to some extent; some more readily than others.

The concepts both helped identify influences and fostered a deeper understanding of

influences. A full discussion of the value of the HAMTS in particular, and the value of a

MTS in general, is offered in the following chapter (See Section 6.2.5).

Table 41 Concepts from the Human Agency Multi-Theory Scaffold (HAMTS) Theory from Human Agency MTS Agency theory concepts used in

analysis of empirical data Meta-Theory: Structuration Theory (Giddens, 1984)

Modalities 1. Interpretive scheme 2. Facility 3. Norm

Critical Theory: Social Theory of Gender Theory (Connell, 1987)

4. Power

Middle-range Theory: Social Cognitive Theory (Bandura, 1986, 1997, 1989b, 2001)

Environment mechanisms 5. Vicarious observation 6. Feedback and reinforcement 7. Scaffolding

Individual mechanisms Self-efficacy

8. Outcome expectations 9. Disposition 10. Self-reflectivity

Operationalised: Social Cognitive Career Theory (Lent et al.,1994)

11. Self-efficacy

5.3.1 META-THEORY: STRUCTURATIONTHEORY (ST) Three concepts from Giddens’ (1984) stratification model of action (a component

of Structuration Theory) were used to re-analyse the empirical data: 1) interpretive

scheme, 2) facility, and 3) norms (as illustrated in Table 41). These three modalities

connect the individual’s actions (communication, power, and sanctioning of conduct) to

the societal structure (signification, domination, and legitimation). Data analysis sought

to identify if the three modalities were evident in the empirical data, or could be used to

explain the data. Table 42 presents examples of matches between the empirical data and

the modalities. Following are several further examples of how modalities can help

explain the influences on women’s participation.

Table 42 Empirical evidence of Giddens’ concept of modalities ST modality An example in the empirical evidence

1 Interpretive Scheme

The modality suggests that stereotypes are communicated in the media or in person and signify their capacity to the person. In the empirical data, participants’ comments that “‘if everyone says” that women are not capable (M3) are suggestive of there being what Giddens calls ‘common stocks of knowledge’.

2 Facility The modality suggests that access to resources is influenced by power relationships. In the empirical data, participants’ comments suggested that access to software could empower an individual; however, many of these resources were targeted towards males.

3 Norms The modality suggests that women face social norms that they are expected to follow. In the empirical data, participants’ comments suggested that they recognised norms such as “It's not expected that the girls are as interested in games as the guys” (G2).

1. INTERPRETIVE SCHEME

Aspects of the empirical evidence can be explained using Giddens’ modality of

‘Interpretive scheme’ (Communication< Interpretive scheme >Signification). The

Interpretive modality draws attention to structures of “signification” (language and other

symbolic codes) to produce/reproduce “communication” (meaningful and understandable

interactions) (Giddens, 1979, 1984). There was evidence that the empirical data

(participants’ accounts and secondary data) can be understood by drawing on Giddens

concept of ‘Interpretive scheme’. There were two specific examples: 1) the signification

of women’s stereotypical capacities presented in books and on the internet, and 2) the

non-signification of gender through clothing in the workplace.

1) Participants made statements in reference to stereotypes surrounding women’s

proficiency with technology: “If everyone tells you the same thing then you will think that

is the truth” (M3). These statements were interpreted by the researcher as representing

participants’ awareness of what Giddens (1984) referred to as ‘common or symbolic

knowledge’. One participant suggested that stereotypes presented by mediums such as

books–particularly the title ‘Why women can't read maps’ (Pease & Pease, 1998) –

perpetuate the notion of women’s limitations in working with technology.

Secondary evidence also suggests that resources within the environment may

signify they are not intended for women. For example, software on the internet is

presented alongside links to live sex shows and porn (See Appendix 17, Item 20).

Giddens (1984) proposed that shared stocks of knowledge, communicated through

language and symbolic artefacts (for example, books), can signify to an individual their

capacity. Books and the internet are examples of these ‘shared stocks of knowledge’,

where sexist material can signify to women certain social beliefs about their capacity to

work with technology.

2) A further empirical example of signification occurs in the clothing typically

worn by workers in the DCI; this is casual attire, such as “super daggy, jeans and t-shirt”

(S1). The male freelancer (S1) noted that the casualness of the workplace attire contrasts

with that of corporate workplace environments: “The image thing in terms of a corporate

workplace environment isn’t there” (S1). The freelancer also noted that clothing does not

signify capability in the DCI workplace: “It doesn’t matter what you look like or what

you’re wearing. You do what you do and you’re rewarded for the product that you make

and if you can make a game that’s successful”. Similarly, a female participants viewed

casual workplace attire as being a positive characteristic of the environment because it

minimised the gender stereotypes she experienced. As the DCI industry allowed

workers to wear whatever they wanted, this meant that you could “be comfortable all day

and you can express yourself”. She continued: “I can’t imagine working somewhere

where you’re expected to wear a skirt, here you can just be yourself all the time and not

worry about being something different” (G3). Casual attire did not ‘signify’ a person’s

gender and, consequently, women felt that they faced fewer gender stereotypes.

Participants linked the wearing of clothing typically associated with women (for

example, skirts) to somehow constraining their individual capabilities, because it

signalled their gender and the stereotypes linked to that gender: “If I didn't understand

something I felt it was because I was wearing a skirt “(M4). Giddens (1993) noted that

gender identification manifests through clothing and gives an example for the education

context, where “regulations that compel girls to wear dresses or skirts in school form one

of the most obvious ways in which gender typing occurs” (p. 43). Thus, Giddens’

modality of ‘Interpretative Scheme’ helps foster an understanding of influences on

2. FACILITY

‘Facility’ (Power<Facility>Domination). The Facility modality draws attention to the

person’s capability to exercise power at the level of concrete practice by accessing

domination structures, which include allocative or authoritative resources (Giddens,

1979, 1984). There was evidence that the empirical data (participants’ accounts and

secondary data) can be understood by drawing on Giddens concept of ‘Facility’.

Although power was not mentioned to any great extent by the female DCI

workers, the researcher deduced three specific examples where power could be at play:

1) the exclusion of women by men in workplace settings (leading to a sense of

powerlessness), 2) the physical layout of workplace settings (which can contribute to

mitigating this exclusion), and 3) access to software and hardware resources (which can

influence women’s participation).

Women described situations where it could be deduced that they felt less than

empowered. For example, when male colleagues conducted work-related meetings in

strip clubs, these practices left women feeling excluded because they felt denied access

to information. In contrast, the physical layout of workplaces was seen as a positive

aspect as it facilitated openness of communication. Giddens (1988) identified that

physical settings are connected to power (pp.286-298). Participants did not identify

resources as being associated with power. However, they did note that access to

resources–such as industry websites, magazines, computers and mentors was an

important influence because it gave them access to information. For Giddens, access to

resources is an aspect of power. Resources such as magazines, computers and software

may be considered as examples of what Giddens (1984) referred to as ‘allocative

resources’, and a mentor as an example of ‘authoriative resources’.80 Thus, Giddens’

modality of ‘Facility’ helps foster an understanding of influences on women’s

participation.

3. NORMS

‘Norm’ (Sanction<Norm>Legitimation). The Norm modality draws attention to how a

person may sanction others by drawing on legitimation structures (moral orders

associated with laws and religion) (Giddens, 1979, 1984). Findings suggest that women’s

participation is influenced by social norms perpetuating inaccurate social stereotypes

regarding women’s capacity for technology. In addition, women face sanctions if they

transgress social norms.

Participants noted that the norms regarding women’s capacity for technology

could limit their participation in the DCI. One participant notes: “It's not expected that

the girls are as interested in games as the guys…I think it's just a general assumption”

(G2). Women appeared to be aware of these social norms and faced the challenge of

reconciling the norms against their own personal experiences: “I think there was a barrier

there in my head. You know programming is sort of seen as you know a boy’s thing ... I

think there’s a message that comes down the chain that you know that girls just can’t do

that stuff; you know they don’t have the mental capacity and all that sort of stuff” (M5).

Another participant noted that an absence of social norms fostered their participation: “I

felt comfortable interacting with guys at an early age ...perhaps I didn’t get conditioned

that way to have separation between girls and boys because there weren't any” (G6).

Another noted the role the individual plays in challenging those norms: “Even if there's a

tendency, like if most women don’t like programming doesn’t mean that all women don’t.

There’s a lot of factors that can add up to what they like. I like programming but I only

know that because I got into it because I liked math” (G3).

Participants noted the norms that suggested women were best suited to

administrative and communication-based tasks in contrast to technical. The social norm

was that: “Women are better at certain things and men tend to be better at certain

things” (G2). Comments from a stakeholder also suggested there were norms regarding

women’s capacities to work in certain occupational roles: “It’s really around the kind of

segregations around skills around organisation, and project management, and following

80 1) Authoritative resources (which derive from the coordination of the activity of human agents and the capability to generate command over other people); and 2) Allocative resources (which stem from control of material products or aspects of the natural world, or the capabilities to generate command over material objects) (Giddens, 1984)

through and attention to detail that women excel in this organisation at. The obvious

ones are communication...” (S3). However, the stakeholder’s comments also suggested

that people’s actual experiences can dispel inaccurate gender stereotypes. In reference to

the norm that women are better communicators he stated: “But I think that's a bit a cliché

the communication one [...] it’s not to say that is always the case” (S3).

Although one participant described feeling “obligated” to social norms (G2),

others felt they could challenge certain norms surrounding occupational practices. For

example, although there was an expectation that team members would work back late,

several participants took the action of not doing so, as they believed it was not required.

Any self-sanction, or sanction from colleagues, was minimised “as long as you’re getting

the work done” (M6):

... it's very easy to allow yourself to feel pushed into certain things. Like I turn up at 8

and leave at five and there's always people that are staying late. But if I finish my work

I go and I think people tend to see everyone else and they feel bad and I think that's a

part of what fosters this unhealthy work life balance is where people feel they have to

even though they don't. (G4)

Women appeared to face sanctions if they transgressed gender-related norms.

One participant identified that women may face some form of retribution from male

colleagues if they did not act the way the norms suggested women should: “It’s a feeling,

yeah, that I got just from the vibes I got” from the “guys” that they “don’t like that strong

female and I think that’s not just multimedia. I think it comes in a lot of different areas

where that strong female has an issue” (M7). Women faced different social norms or

‘rules’ than men: “I think the rules are a little bit different for women than for men” (G4).

Thus, findings suggest that women face norms as a result of their gender and

although women could challenge these norms, they risked facing sanctions for doing so.

Thus, Giddens’ modality of ‘Norms’ can help explain the empirical data and foster an

understanding of influences on women’s participation.

5.3.2 CRITICAL THEORY: SOCIAL THEORY OF GENDER (STG) Empirical data was also re-analysed using Connell’s (1987) Social Theory of

Gender (STG), with specific regard to his conceptualisation of power. However, there

was limited evidence that the empirical data (participants’ accounts and secondary data)

could be understood by drawing on Connell’s notion of power. Analysis revealed that

power was not a concept often raised by participants as being an influence on their

participation. Only two participants explicitly mentioned the term ‘power’ and this was

in reference to difficulties they experienced when working with other women, rather than

men. In contrast to Connell’s critical theory, which draws attention to inherent power

relationships between men and women, one participant described the power play that

occurred between women:

When you have two head strong women and there's that power play, you know power

roles and I don't think women really negotiate power very well in general” ...“I think

women sometimes have a disregard for each other’s power because they want the

same thing, OK, and they are not really willing to acquiesce to the other persons, umm

you know position or authority. (M5)

However, power could be inferred from the data. For example, several

participants highlighted that once they felt they had gained experiences and skills, they

were in a stronger position to negotiate their salaries. One stakeholder made a direct

reference to ‘power’ in reference to certain roles in the industry; for example, she saw

programming as providing individuals with power: “It's somewhere where you can find

power. Programming is power” (S2). She described that the PC (personal computer) is

not only a toy, but a young male’s “world of strength” (S2). She implied this was not often

acknowledged: “If they ever realised how powerful they could be. Well society would

collapse and they’d run amok, wouldn’t they?” These findings suggest that it is a

person’s own sense of personal power, rather than power play between the genders, that

is an influence on their participation.

5.3.3 MIDDLE -RANGE THEORY: SOCIAL COGNITIVE THEORY (SCT) Data was further re-analysed using several concepts from Bandura’s (1986)

Social Cognitive Theory (SCT). As Table 43 indicates, these concepts include the four

inter-related cognitive mechanisms of the person: 1) self-efficacy, 2) outcome

expectations, 3) disposition, and 4) self-reflectivity. Three further mechanisms that

involve the environment were utilised and these were: 1) vicarious observation, 2)

feedback and reinforcement, and 3) scaffolding. There were many instances of

resonance between SCT concepts and the empirical data. Table 43 provides several

examples of matches between the empirical data and the person mechanisms; these are

then further detailed. Similarly, Table 44 presents several examples of matches between

the empirical data and the environmental mechanisms.

Table 43 Empirical evidence of Bandura’s person mechanisms Person Mechanism

As evidenced in empirical data by participant quote

1 self-efficacy

-“Umm I think you have to have self-confidence I think that’s the only way and

have a goal” (M4).

-“if you don’t have the industry experience in a proper employment setting. It

sort of eats into your professional confidence in your own personal ability and

stuff” (M5).

-“for me confidence came from experience” (M4).

2 outcome expectations

-“I like that it will get rated highly by the reviewers so I don't want to work on a

cookie cutter whatever game that will just get thrown off the shelf”(G3).

-“I do want to work on a game that I think is going to be fun and cool, and my

little brother will like” (G3).

-“I mean, yes, you could definitely do the job competently but would you find it

rewarding, would you really want to be there” (S1).

3 disposition

-“I think it does attract a certain type of female sort of personality type” (G6).

- “Going into an all women’s class I thought I had a better chance of getting

higher grades” ...“because I had always had to put up with guys who were

'always' better for whatever reason” ,“it's probably just an ego boost”. I'm

someone who doesn't like to fail” (M2).

4 self- reflectivity

-“...wow when you think about it that’s interesting” (G2).

-“Even though years after starting in the workplace retrospectively I can go, I

was still really lacking confidence, it was building up… a lot of the experiences

since working professionally have been really positive” (M4).

PERSON DRIVEN PROCESSES

1) SELF-EFFICACY

The findings suggest that a woman’s developing sense of confidence or self-

efficacy towards technology is integral to her participation. Although participants did not

use the term ‘self-efficacy’, they often used the term ‘confidence’: “They will succeed…if

they have self-confidence” (M4). ‘Confidence’ is interpreted here as a colloquial form of

self-efficacy, as many of the references to confidence were in relation to skill

development. Confidence does not necessarily specify what the strength of belief is

about, whereas self-efficacy includes both an affirmation of a capability level and the

strength of that belief (Bandura, 1997). Bandura (1986) defines self-efficacy as

“people’s judgments of their capabilities to organize and execute courses of action

required to attain designated types of performances” (p. 391).

The findings suggest that self-efficacy manifests when participants have the

opportunity to develop skills that foster their ‘confidence’. As people practice and

develop their mastery of a skill, they become more confident and have a belief in their

ability and capacity. Several participants recounted how surprised they felt when they

entered the workforce and realised “it wasn’t as hard as they thought”, and that “they just

picked it up” (in reference to computing skills): “This role has given me the chance to

develop my skills further and embrace new professional opportunities that would not

have been possible before. This has provided me with an enormous amount of personal

satisfaction and confidence in my abilities” (M5). Thus, participants’ confidence was

linked to their experience: “For me confidence came from experience” (M4)...“once

you've got the experience it doesn’t matter, people want you, you’re very valuable” (G5).

The findings also suggest that self-efficacy manifests over a lifespan perspective,

where early childhood and education mastery experiences foster self-efficacy towards

technology later in life. Ongoing mastery opportunities were required, particularly as

“technology moves very fast and it takes time to master every version of something that

comes through” (G8). This ongoing need to maintain skills was seen as both an enabling

and a constraining influence by one participant. It was enabling as it provided her with a

motivating challenge; however, it was also a constraint as it involved an ongoing

commitment and effort to keep up with changing technology: “In the early days I was

really excited when a new version came out and I wanted to know everything that it did

and I wanted to know all the little things that were going to happen. But after I did that a

couple of times I thought, yeah, I'm not going to be doing this...” (G8).

Having both the skills and confidence could influence interactions such as salary

negotiations. One participant recounted that women wouldn’t ask for certain salaries

(comparable to men’s) because “they didn't have the confidence to ask for it”, whereas

“the guys have the confidence to tell the boss to fuck off” (M2). This was an observation

the participant made of “Not only myself but most other girls” who she “knew had trouble

getting the salary that they were worth. Not because they didn't want it, but because

they didn't have the confidence to ask for it” (M2).

Secondary sources further support the notion that ‘confidence’ is central to

women’s successful participation in the industry. For example, several articles indicate

that women may lack confidence. For example, Appendix 17, Item 11 reports that an

American games industry panel suggests that “women don't always recognise or even

have confidence in their ability to be games designers”. Industry initiatives aim to help

women increase their confidence. For example, an email from an industry group for

women in IT in the UK asks: Do you lack confidence? If so, it offers workshops to

address this (See Appendix 17, Item 17). Being in the gendered minority could diminish

women’s confidence “because you doubt yourself you know, naturally, you doubt your

confidence because there aren’t any other females in the industry that I can look up to”

(G5). Additionally, women were seen as not possessing this confidence; in part, because

of not wanting to appear confrontational:

... that confidence is going to help a lot of people get jobs. A lot of women don’t like

direct confrontation [...] we're afraid of offending people a lot more and we're not so

aggressive and I think that comes down to confidence as well and experience. (G5)

2) OUTCOME EXPECTATION

Bandura recognised that individuals have outcome expectations, where they may

ask the question: ‘If I do x, what will happen?’ Outcome expectations are often driven by

rewards (intrinsically or extrinsically driven), or perceived rewards. According to

Bandura, such rewards include physical and monetary rewards, social approval, and self-

satisfaction (Lent et al., 1994, p.83).

Participants identified that gaining skills offered an intrinsic reward, which then

could serve as a further motivator: “Once you start learning and getting results and you

actually make a program you can get a reward to say wow this is fun, I want to make it

more complex” (G5). For some participants, outcome expectations involved perceived

financial reward. One participant (who had a Computer Science degree) reflected that

monetary reward had been a motivating factor for her choosing to pursue games

development in contrast to other areas: “I think that was probably my main deciding factor

was all that hype about the money and I thought I'm going to get some money”; there

was that “perception back then of being worth so much money” (G3). Secondary sources

of evidence, such as the Women in the IT Industry survey, also indicate that possible

motivators may be money (Appendix 17, Item 41).

Other rewards that sustained participants’ motivation included the ability to be

creative. Some participants valued the creative rewards higher than the financial:

It’s not an industry you make a lot of money in. It’s about the artwork [...] for me it’s

about the game that I make[…]I’d love to go and work in Moldavia for 6 months for no

money on one of these crazy titles that they’re producing […] they look fantastic. (G7)

Where outcome expectations reflect a person’s beliefs of what may happen, goals

represent what an individual would like to do. One participant suggested that women

may have clearer goals than men: “I don’t know if it’s a general male or female thing but

women seem to have more of a future sort of plan” (G7). Participants’ accounts

suggested that they pursued situations where they could meet their goals: “At school I

wanted to get good marks so I always took the courses I would excel at and so in

university I ended up with a computer science focus” (G3). One particular participant

described the ‘safe’ (M2) all girls education environment she experienced as a student as

having fostered her participation75F

81. While she used the term ‘safe’, the participant did

not feel threatened in any way, but wanted an environment where she could meet her

goals: “I guess my understanding of safe is more about achieving my goals not

necessarily being petrified of the boys or anything like that” (M2).

Women’s goals or outcome expectations could, however, be constrained by the

industry characteristics and gender stereotypes that they face. This is exemplified in the

practice of funnelling women towards administrative roles, which may not meet the

individual’s needs in light of their skills and aptitude:

I want to see that I can progress and the women that I’ve met are slightly a little bit

more career oriented, they like to see that they can progress, not necessarily in the,

dare I say, administrative line. I mean you don’t want to study for X-amount of years to

become the office manager. (G7)

3) DISPOSITION

In Social Cognitive Theory (SCT), an individual’s disposition is a dynamic,

multifaceted belief system and includes their reactiveness, self-regulation, and response

(coping strategies and attitude based on experience) (Bandura, 1999b). An individual’s

disposition may mean that they respond differently to a situation or circumstance than

others.

One participant described a situation where she and fellow female colleagues

were bullied by a male manager. However, her individual response differed to that of one

particular colleague: “One of the managers there who was an outrageous bully, like just

insane and I just started not tolerating it. Initially I used to get really upset and not really

understanding what was going on” (M4). She noted that the fellow worker had a “nervous

breakdown” because of the stress and, consequently, left not only the organisation but the

industry. In contrast, the participant responded differently; the negative conditions

motivated her to pursue job opportunities elsewhere. This action led to her subsequent

employment in a public organisation, which she perceived as providing better conditions.

Thus, the individual differences amongst women may involve their unique dispositions.

4) SELF-REFLECTION

For Bandura (1986) the capability that is most “distinctly human” (p. 21) is that

of self-reflection. Through reflective self-consciousness, people evaluate their

motivation, value commitments, and maintain meaning of life pursuits (Bandura, 2003,

p. 168). Findings indicate that participants were self-reflective towards their

participation.

81 She was participating in a Diploma of Multimedia for Women (with an all-female student intake).

Participants’ reflectivity can be interpreted from comments such as:

“Retrospectively I can go, oh, I was still really lacking confidence, it was building up”

(M4). One participant described a strategy she had actively employed to return to work

after having her first child. The key requirement for her was to reduce work pressures

and balance family life. She described how she ‘took stock’ when making the decision to

delay pursuing working on Triple A titles that were seen as crucial to establishing her

occupational credibility and job opportunities: “So when he’s a little bit older then I’ll sort

of hit some AAA-titles, so it’s a good time to just sort of take stock” (G7).

Other participants exhibited reflectivity during the interview. For example, when

asked how many women worked at her organisation, one participant noted that they had

thirty women, implying this was a healthy figure. When asked how many were in

technical roles (in contrast to administrative) she identified “under ten maybe 5 or 6”

(G2). It was the interview situation that made her consider this difference and she noted:

“Wow, when you think about it that’s interesting” (G2).

ENVIRONMENT DRIVEN PROCESSES

In addition to the four cognitive mechanisms that the person possesses, findings

indicate that aspects of the participants’ participation could be explained by three further

mechanisms that Bandura identified. These mechanisms stem from the environment the

individual is positioned within, rather than from the individual themselves. Accordingly,

empirical data was coded for the following three mechanisms:

1. Reinforcement or feedback

2. Vicarious learning experience

3. Scaffolding of skills and knowledge

Table 44 provides an example of empirical data matching the environment mechanisms that Bandura’s SCT draws attention to.

Table 44 Empirical evidence of Bandura’s environmental mechanisms Environmental mechanism

As evidenced in the data

1 Reinforcement or Feedback

-“It wouldn't surprise me at all if it is more about how they are

treated in the workplace whether like, lack of feedback” (M2)

-“... she didn’t like the criticism that she was getting and I think

that’s probably because the criticism, like that was all she was

getting from the environment, just criticism and no positive feedback

but no neutral feedback, no general chit chat, none of those” (M8)

2 Vicarious experience

-“There isn't an image of a hot female programmer out there that

people want to emulate or aspire to” (G10).

3 Mastery opportunities/ scaffolding of skills and knowledge

-“I want to work with someone who's had more experience as well

that I can learn from” (M6). -“Career opportunities arise from

experience through this experience we seek out or given the

opportunity to move onto something else and be mentored” (G2).

1) REINFORCEMENT OR FEEDBACK

Bandura’s SCT suggests that feedback can be received through ‘verbal

persuasion’ from other people (Bussey & Bandura, 1999). Receiving feedback on

performance accomplishments will improve a person's self-efficacy and help a person

consider if their efforts and goals are feasible and realistic.

Participants identified that receiving feedback was a positive influence on their

participation as it facilitated their sense of feeling valued as members of the industry and

increased their confidence: “Management say good things and that was an enormous

umm [...] confidence booster, so yeh I kinda thought this is not so bad” (M5). Feedback

came from a number of sources, including employers, peers, and even clients. Employers

could provide feedback, informally or formally, such as during a performance review:

“In my review I had a sense that I was being treated completely fairly and umm that I was

being supported” (G3). One participant thought back to her first day at the DCI

organisation she was employed at, and described how a colleague was receiving praise

for a presentation she had given: “Just the support and congratulations, I don't know, I

just thought that was like wow, whether it was true or not I don't care but they came out

and said ‘done good”. She added: “I want some of that” (M2). Another participant

described how a male colleague she had worked with previously recommended her for a

job in the company he worked in. His recommendation seemed important for her self-

confidence: “For me that was sort of a bit of an achievement […] I actually felt confident

probably for the first time in my whole career that I could actually do this work” (M5).

Another participant, whose role involves some Flash76F

82 software programming,

and who found programming fun, stated that “women were not good at programming”

(M2). When asked if that was nature or nurture, her reply indicated a level of reflexivity

about the influence of socialisation: “I'm not sure, whether it's just because everyone

says that, then we grow up thinking that, so we sabotage ourselves” (M2). Although she

had used the word ‘we’, it is unclear to what point she was self-stereotyping, as

previously she had identified herself as not being like other ‘girls’. Her comments

suggested that social messages may reinforce inaccurate stereotypes regarding women’s

capacity for technology-related careers. This is not dissimilar to Giddens' modality of

‘Norms’, which recognises that individuals face sanctions from others.

2) VICARIOUS OBSERVATION

Bandura’s SCT suggests that learning from others through vicarious observation

can foster the person’s cognitive mechanisms identified earlier. Learning from

observation is enhanced when the people involved share similarities, such as gender

(Bussey & Bandura, 1999); for example, female workers observing female role models.

Participants indicated that they learnt skills and social practices by observing

other people; both in their immediate context and via mediated forms such as television.

Participants utilised terms presented in Bandura’s theory, such as ‘role models’, to

describe these people. A majority of the women interviewed, in both multimedia and

games production, believed that the lack of suitable female role models was a key

negative influence on women’s participation: “I reckon there are some positive role

models out there you kinda just have to sort of bump into them along the way. But there's

no real, like again it comes back to the public face” (M5). Only one of the women

interviewed could identify a female role model who was a working mother. There was a

paucity of female role models who were programmers: “I’ve only ever heard of one other

female [programmer] and this is a story someone else was relaying to me about a

country they used to live in” (G2); and “There isn't an image of a hot female programmer

out there that people want to emulate or aspire to” (G4).

In contrast, several secondary sources indicate that there are role models, at least

internationally. For example, the games industry website Gamasutra identifies the top 20

role models, and the Edge magazine lists the top 100 (Appendix 17, Item 23). Events

such as the 2010 Women in Gaming Awards Luncheon (Appendix 17, Item 13) aim to

recognize “the accomplishments of all women in an industry currently dominated by

men”. However, although sponsored by well-known industry groups (in this case

82 Flash is a software package from Adobe that utilizes Actionscript. ActionScript is an ECMAScript-based programming language.

Microsoft), the event is described online as an ‘intimate’ event. This suggests that

although role models exist, they may not be publicised widely.

There are also few resources which promote female DCI role models in

Australia. For example, the Women in Games in Australia website contains very little

information overall, and does not identify any specific women working in the industry

(perhaps other than the author of the site) (See Appendix 17, Item 18). One particular

role modelling event the current study identified is a Wired Women talk held as a part of

the Game On conference (Brisbane, 2009).77F

83 Online comments from a blog reporting

this event indicate that it provided suitable role models (See Appendix 17, Item 50); so

much so, that an attendee blogged how motivating the event had been:

I enjoyed the talk immensely. I liked that it was more focused on the career side of the gaming industry and how it is open to women - I was pretty inspired to run out of the room and go finish my IT course. (Appendix 17, Item 43)

3) SCAFFOLDING

Bandura’s SCT suggests scaffolding involves an individual learning skills and

practices by working with and being guided by other people (or entities).Learning in the

work setting often occurs through mentors. Bussey and Bandura (1999) suggested that

same sex role models enhance scaffolding experiences, increasing role identification and

even rewarding motivation. However, findings reveal there are few female mentors in

the participants’ immediate work environment: “They're all hims...there's no hers” (G1).

A games worker further noted:

There’s not a lot of role models and so that if you were to go into one of those roles you

wouldn’t necessarily have someone, another woman, that you could look up to or

mentor, who would mentor you and who you would ask for advice. That sort of stands

out to me. (M8)

Scaffolding involves mastery opportunities, which could be facilitated by other

people or by access to resources. For example, access to learning resources in their

childhood helped one participant gain programming skills: “My dad would get

commodore 64 magazines and so I'd just copy out programs” (G2). One stakeholder, an

educator, identified the need for such scaffolding when teaching her students

programming within a university degree focusing on games development. She

highlighted that, because learning programming presents challenges for the diverse range

of students who come from design, animation and writing backgrounds, “you need all of

the scaffolding that you can to get into it” (S2). She went on to identify the limitations in

83 The event was held at the State Library of Queensland. The researcher both attended this event and searched for online comments (on the internet) following the event.

how programming was taught: “When you do get exercises they’re the same ones that

used to give us in maths. You know those really weird ones like one man digs a hole in a

field and it takes him six hours” (S2). She made the point that boys persevere with

learning coding “because you get a lot of boys who actually possibly already kind of

realise that that’s where their friends are” (S2). She suggested boys have mastery

opportunities early on, before they enter formal education pathways: “Some of them

come in already kind of implicated into that sort of field. They know where they want to

be because they know before, they’ve seen it” (S2).

5.3.4 OPERATIONALISED THEORY: SOCIAL COGNITIVE CAREER THEORY (SCCT)Empirical data was re-analysed using only one concept from Lent et al.’s (1994)

Social Cognitive Career Theory (SCCT)–the concept of self-efficacy. Self-efficacy (or

more informally, ‘confidence’ and skill development) featured strongly in participants’

accounts. SCCT offered theoretical triangulation with Bandura’s SCT regarding self-

efficacy. Furthermore, Lent et al. (1994) also presented 12 propositions regarding their

theory; several involve person characteristics such as values (p. 91) and problem- solving

(p. 99), and others include environmental influences such as social persuasion (p. 102).

As will be discussed in 6.2.5 in the next chapter, in later stages of analysis and

theorisation, Lent et al.’s (1994) work sensitised the researcher to other possible

influences.

5.4 FRAMEWORK 3: ONTOLOGICAL (CRITICAL REALISM)

Chapter 3 proposed that an ontological framework, based on Critical Realism,

could help guide data analysis. Critical Realism requires the researcher to identify the

causal effects at play; that is, the underlying mechanisms. Accordingly, empirical data

was analysed using Bhaskar’s ‘three domains of reality’: 1) empirical, 2) actual, and 3)

Analysis revealed the participants’ experiences, specific events, and importantly,

the underlying mechanisms that manifest in the interaction of the environment and

person. As Table 45 illustrates, earlier analysis of participants’ experiences using

Framework 1 provided an insight into the first two domains of reality, ‘empirical’ and

‘actual’. Therefore, there was no further analysis of these two domains. Similarly,

Framework 2 had sensitised the researcher to mechanisms involved in interaction, mostly

through an abductive logic. Re-analysis of the empirical data, by employing retroductive

logic, led to the identification of 10 mechanisms, which are presented as the Five Acts of

Agency (FAA) model.

As discussed in the following chapter, the combination of the insights from the

analysis of data using the three frameworks (proposed in Chapter 3) led to the

conceptualisation of agent-driven mechanisms. The findings here begin to report on the

theorisation of the phenomenon of women’s participation; however, detailed discussion

is reserved for the following chapter.

Table 45 How the three frameworks illuminated the three ‘domains of reality’ Domain Real Actual Empirical Findings

Experiences (The real and the actual that is experienced by individuals)

x x x Empirical experiences of participants were illuminated by Framework 1.

Events (The domain of events)

x x Framework 1 also drew attention to 10 specific events of interaction. These events of interaction provide an insight into the ‘actual’ domain.

Mechanisms (The domain of causal tendencies)

x Framework 2 sensitised the researcher to mechanisms. Framework 3 explicitly focuses on abstracting mechanisms

5.4.1 AGENT-DRIVEN MECHANISMS

The findings reveal 10 mechanisms involved in the interaction between a Person

(P) and their Environment (E). These mechanisms are driven by the person, in contrast to

those that arise from the environment. When women harness these mechanisms, it can

result in actions that foster their participation. As Table 46 presents, these agent-driven

mechanisms are: 1) accessing, 2) imagining, 3) doing, 4) belonging, 5) sharing, 6)

problem-solving, 7) transforming, 8) emotion, 9) ethical and 10) being. These

mechanisms are organised into five categories of ‘acts of agency’, which are: 1) enabled,

2) connected, 3) collaborative, 4) creative, and 5) human. These ‘acts of agency’, and

the mechanisms within them, are presented as a model of the Five Acts of Agency (FAA).

Table 46 The Five Acts of Agency: The underlying agent-driven mechanisms

Five Acts of Agency

1. Enabled

Access to core tools and

resources such as technology, people, and information

2. Connected

The beliefs and motivations of

the agent, including self-

efficacy

3. Collaborative

The social relationships that

foster involvement

4. Creative

Problem-solving, or the power to

transform

5. Human

Human-only traits of emotions

and ethics

Agent-driven mechanisms

1. Accessing (Technology, Information, People)

2. Imagining3. Doing

4. Belonging5. Sharing

6. Problem- solving 7. Transforming

8. Emotion 9. Ethical 10. Being

AN ACT OF AGENCY

There are three components to the ‘act of agency’ concept (as illustrated in

Figure 12):

1. the agent-driven act;

2. the specific mechanisms involved in the particular act;

3. the outcome or consequence of that act.

As the following chapter discusses, the ‘agent-driven act’ (Point 1, Figure 12) is a

high-level concept combining Human Agency theory and Critical Realism. The ‘agent-

driven mechanisms’ (Point 2) include both those mechanisms that the HAMTS sensitised

the researcher to, and those that emerged from the retroductive process; in both instances,

these are supported by empirical data. The ‘outcome of the act’ (Point 3) requires

consideration of the consequence of these acts and mechanisms, which in the current

case study, is women’s participation.

Figure 13: Components of an ‘Act of Agency’

THE FIVE ACTS OF AGENCY (FAA)

Each of the agent-driven mechanisms within each of the five acts (or categories)

is illustrated with concrete empirical research data (Figures 13-18). Table 47 provides an

exemplar for each of the agent-driven mechanisms. As a Critical Realist approach

requires that theorisation remains linked to the concrete data, Chapter 6 presents further

empirical data to support the researcher’s conceptualisation of mechanisms.

1. Agent- driven act

3. Outcome of the act 2. Agent-driven mechanisms involved in the act

Table 47 Five Acts of Agency: Empirical evidence of agent-driven mechanisms Act of Agency

Example of mechanisms in the empirical evidence

Illustrative quote

Enabled

Access (Technology, Information, People)

Access: Access to resources such as hardware, software, and employment pathways via networking

“…unless you’re really well

networked it's quite difficult to get

a leg in, in this city” …“you have to

find a way to become involved and

there are barriers to that because

it's not something highly visible”

(M5). “…to find that environment I

think that was the hardest thing”

Connected

Imagining, Doing

Imagining: Aspiration and motivation scaffolded by role models and goals Doing: Developing skills and confidence in the domain area

“there isn’t an image of a hot

female programmer out there that

people want to emulate or aspire

to” (G4).

…“it just came very easy to me.

That made me start thinking

maybe I can actually go a bit

further with this” (M4).

Collaborative

Belonging, Sharing

Belonging: Norm knowing and awareness of sanctions Sharing: A sense of contribution and feeling valued

“…there's a lot of reward to be

gained from collaborating with

people. People ask me questions

and I ask them questions about

stuff which shares skills and

knowledge” (M5).

Creative

Problem- solving, Transforming

Problem-solving: Enjoying the challenge of learning software, meeting client briefs; intrinsic rewards of solving problems Transforming: Intrinsic rewards of creating new objects or practices

“The things that I didn’t enjoy

about it was that it’s not a creative

industry and that was what I

wanted” (G7).

“I like challenging projects that

make me think and bring about

new ideas, finding new ways to

resolve them, improving myself

over time” (M7).

Emotion, Ethical, Being

Emotion: for example, passion, guilt, risk anxiety Ethical: moral congruence, types of products Being: personal reflection

“It's a cycle in that you know

everyone is passionate about

work to put in these hours” (G6).

1) ENABLED ‘ACT OF AGENCY’The Enabled Act of Agency requires the person to employ the mechanism of

‘accessing’ resources such as information, technology and people.

Figure 14: The Enabled ‘Act of Agency’

INFORMATION- The mechanism of accessing was evident when participants

indicated that access to information was an influence on their participation. A lack of

information was seen as limiting the choices participants made: “I don't think it was ever

something that I was interested in because I didn't know enough about it but now I know

it I don’t think there's anything else I’d prefer to do” (G1). The actions women took were

to purposively seek out information, such as job postings on game development websites

(such as gamedev.net). Another participant suggested that the education system had not

provided her with adequate information regarding career choices: “School didn't give us

that information I went to company websites...It's hard to really say definitively girls are

being discouraged or whether there's just not enough information…it never occurred to

me in high school there was no information perhaps ...I didn’t take IT when I was in high

school and I regret that now” (G2).

TECHNOLOGY- The mechanism of accessing was evident when participants

indicated that access to technology, in particular computers, was an influence on their

participation. Access to free software was seen as integral to developing skills that may

lead to employment: “It’s interesting because there are new ways to get a job… there’s

free online engines so these kids are actually making them at home” (G7). Technology

resources also offered flexibility to work from home, which may be relevant for working

parents: “Flexibility is ...I might be able to do some work at home because I’ve got all the

software at home and I’ve got access to work” (G7). Technology resources were

important over their lifespan, including their childhood, education and careers. For

example, a lack of access to computers presented women with challenges during their

education:

When I was doing it, it was even harder because hardly anyone had a computer at

home, you’d have to go to the lab and wait two hours until there's one free and you're

only allowed on for three hours but your lab for that week could take nine hours. (G3)

PEOPLE- The mechanism of accessing was evident when participants indicated

that access to people (both people in the immediate organisation and also across the

industry) was an influence on their participation. Having connections or networks was

integral to women’s entry and continuation in the workforce; however, many positions in

Enabled

Accessing (Information, Technology, People) Participation

the DCI are not widely advertised: I didn't think we advertise jobs that widely. To be

honest most people that come in are people that someone already knows” (G6). Access

is often fostered through social connection based on friendships and professional

networks. The industry was “competitive”, and “unless you’re really well networked it's

quite difficult to get a leg in, in this city” (M5). Also, “having friends who are involved in

the industry is a good way to start” (M5). Connections could occur in unlikely places; for

example:

To get into it it's a definite – networking is really, really important. I got my first job joining

the [IGDA], just going down for casual beers and a barbecue and meeting people who

are in the industry” […] “you have to meet the right people who are making the big

games and the decisions. (S1)

A participant described feeling ostracized because she had been moved into a

management position where she was actively discouraged from mixing with the

employees she managed, and where most of the others managers in the organisation were

male. Therefore, this participant was pressured to forgo one of her key social links in the

organisation. In response she took action, and “looked to network with my clients, rather

than my work buddies and that really worked” (M4). Later, she started going out with one

of the males who worked in the production area and, through this, “built up a network”. If

there was no social network in the immediate workplace, participants utilised alternatives

such as Facebook and LinkedIn.78F

2) CONNECTED ‘ACT OF AGENCY’

The Connected Act of Agency requires the person employ the mechanisms of

‘imagining’ (or visualising) their participation, and then gaining appropriate skills

through ‘doing’.

Figure 15: The Connected ‘Act of Agency’

IMAGINING - The mechanism of imagining was evident when participants

indicated that being able to visualise or imagine their participation was an influence.

There were several environmental influences that constrained women in doing this,

including the lack of role models, unclear career pathways in the DCI, and unclear

rewards (for example, lack of information regarding salaries). A participant described

how seeing the industry context whilst on a volunteer work placement in a games

84 LinkedIn is a social networking website for professionals.

Connected

Imagining

Doing Participation

development company, helped her connect to the industry. She said that “just the fact that

I was being invited into a games to see what it was like on the inside, that was a huge

thing”, and that “you really need that connection”.

Secondary data sources support the finding that participation may be fostered if

women or (as the example below suggests) young girls imagine themselves in a DCI

role. The following comment is by a blogger reporting on strategies to foster women’s

involvement in games. Visualising or imagining future rewards can be a motivator:

‘Open Days’ for school children so that they can realise this kind of career DOES exist - she believes that is the age that you need to capture someone's imagination for it. (Appendix 17, Item 43) There was a paucity of information regarding careers in the DCI: “Without the

active promotion, it’s not on peoples radars” (G2). There was also inaccurate information:

“ If you haven’t seen it you always think of IT as boring numbers” (G7). Participants

suggested that the lack of accurate information constrained their ability to visualise their

participation:

The major barriers for females entering this industry would be in their own minds. Most

developers are keen to hire women, perhaps even over men, but they don't apply for

the jobs. I think maybe women see it as a men’s industry. (G6)

DOING- The mechanism of doing was evident when participants indicated that

developing skills required by the industry was an influence. When a participant mastered

technical skills, they felt confident in pursuing that career path - they gained a sense of

self-efficacy: “It just came very easy to me. That made me start thinking maybe I can

actually go a bit further with this” (M4). The empirical data suggests that participants

could gain, foster and maintain skills outside of formal training by using “tutorials,

reading articles doing everything I could” (M6).

3) COLLABORATIVE ‘ACT OF AGENCY’

The Collaborative Act of Agency requires the person employ the mechanisms of

having a sense of ‘belonging’ and ‘sharing’ experiences with other agents.

Figure 16: The Collaborative ‘Act of Agency’

BELONGING- The mechanism of belonging was evident when participants

suggested that being able to feel as though they belonged in their environment was an

influence on their participation. Participants valued collaborative and socially orientated

Collaborative

Belonging

Sharing Participation

work environments, where “everyone has input” (G3), and it “feels more just like a group

of friend’s kind of interaction” (G4). Although participants emphasised that they valued

the social aspect of the workplace, they also identified that the male dominated

environment could constrain their ability to make friends: “Maybe it's more of an

acceptable area of interest for the guys than the girls, and more of their peers are in

there, so you know, with all of the social network and all of their friends in there” (G8).

Women found it “frustrating” in regards to “the chance to make friends” (G3). When faced

with male dominated education environments at university, one participant found living

in a female only campus residence helped her make friends whom she could then “draw

on” (G3), even though they did not attend her IT classes. Secondary sources suggest that

a sense of belonging is important in the education context. For example, comments by

the IGDA Women in Games SIG panel (Appendix 17, Item 11) identify that women may

feel as though they don’t belong in games degrees:

Fullerton observed that when young women come to her game design class, they often sit in the back and think that they don'tbelong [emphasis added]. (Appendix 17, Item 11) A need to interact with other people featured highly for participants. One

participant stated that even though she valued a high level of autonomy, where “It's kinda

my way” (M1), she enjoyed working in a team. Another participant identified that

although she enjoyed the flexibility of her freelance work, she needed to interact with

other people, face to face, rather than through mediating technologies: “I just like that sort

of personal interaction […] you sort of get more than just some email” (G8). The DCI,

for the most part, presents a team oriented environment; it is no longer “a guy

programmer playing on his computer coding in his garage”, but rather “quite a close knit

sort of group” (M2) with teams of “forty people for one Playstation game” (G2).

In this team environment, being liked by others was crucial. Participants

described the positives of liking colleagues: “I actually found that I really liked the people

there and I really liked the team... It was a fantastic environment, it was a really, really,

really good environment” (G8). However, even though the participant liked her

colleagues, she felt isolated in some ways because “I was like the only one doing anything

in graphics and multimedia” (G8).

A sense of belonging balanced other influences likely to negatively influence

participation. For example, long hours associated with ‘crunch time’ were seen as being

less of a negative influence as a result of the sense of family in the workplace: “You're

working with such lovely people […] like a big family” (G2). Hence, relationships in the

work environment were important and participants described scenarios that negatively

impacted on their participation as involving “shit work relationships” (M4). A positive

work environment was where people shared experiences. As an employer noted: “In a

work environment you need people who understand you and you can talk to...yeh your

mates with and you can share things and I think that's what a balanced work

environment is” (S3). However, women sometimes faced certain constraints in feeling as

though they belonged as a result of the actions of male colleagues: “It still feels different,

like I feel like they're treading around me” (M6). Conversely, they could also feel

supported by their male colleagues: “I'm lucky that the boys on my team are supportive”

(G1). Practices such as work related meetings being held in pubs and strip clubs can

hinder women’s capacity to collaborate with colleagues. Even in meetings held in the

workplace, women could feel as though they were different–the minority:

I really like the meetings we have because it's like a massive brainstorming session but

then if it's a whole team thing then it should be a whole team thing if you know what I

mean and yeh that's when I do feel like a minority. (G1)

Like often if there’s sort of a whole group of men they won’t necessarily try and include

a woman and she may not feel like – if they’re all talking about football I’m not going to

put up my hand and talk about football because I don’t know anything about that, for

example. (M8)

There were numerous empirical examples of how women could develop a sense

of belonging, for example, through the clothing worn (as previously noted). Belonging

may also occur through speaking the same language, such as technical jargon, and

sharing interests and leisure pursuits, such as playing games. In some cases, a sense of

belonging occurred simply due to the length of time colleagues had worked together:

“I’ve worked with the guys that – who are the artists for the last three or four projects so it

works on a personal level” (G7). For one participant, this sense of belonging balanced out

the limited opportunities for skill development within her (public) organisation: “the

lovely people, that's why I stay” (M2).

Participants identified the importance of a sense of support from other people for

their career progression: “Unfortunately it's really hard to go up from an associate or an

assistant producer to being a producer to take on that complete responsibility without the

support of your company” (G5). This support came from different sources, including a

“mentor or just someone, or in my case my boyfriend who's also on the same team” (G5).

Some women were very active in fostering collaboration, recognising that their career

progression hinged on being perceived as fitting in and accessing opportunities:

I was really aggressive about wanting to fit into the team, about wanting to learn as

much as possible, to compete with the other juniors that were hired at the same time

as me. I know I talked about a collaborative environment but I still like know I'm being

measured. So I always went to lunch with the bosses and we talked about work. (G3)

SHARING- The mechanism of sharing was evident when participants suggested

that being able to share experiences and information, knowledge, and resources with

others through social interaction was crucial for women’s participation. For example,

participants recognised the value of learning from mentors: “I wanted someone to learn

from” (G8). However, there were few mentors, and fewer female mentors. Working alone

was seen as a negative influence: “I was it, so there was no-one there to learn from. So I

was just enjoying the social aspect of it, and doing that job, which was quite challenging

because I was only one doing it” (G8). Sharing was fostered through the availability of

information and resources, which more than often where accessed from outside a work

context; for example, from online communities which offered not just a social value but,

importantly, a professional connection. One participant, a web coder, indicated that

everyone in the web development community relies on each other’s code, such as

JavaScript and CSS: “You borrow very, very heavily. So yeh it seems like everyone works

of each other’s back in this industry” (M5).

4) CREATIVE ‘ACT OF AGENCY’

The Creative Act of Agency requires that the person employ the mechanisms of

‘problem solving’ and ‘transforming’ objects or entities.

Figure 17: The Creative ‘Act of Agency’

PROBLEM SOLVING- The mechanism of problem-solving was evident when

individuals noted that intrinsic rewards, often involving solving challenging problems,

influenced their participation: “It's not a professional environment at all it's really like

getting together with a bunch of friends and solving problems and in that respect it's one

of the brilliant environments to be employed in” (G2). Problems included coding

problems, meeting clients’ briefs or staying up to date with technology.

Problem-solving skills were seen as being integral to certain roles, such as

programming. DCI workers needed to have a “fascination with how things kinda of work”

(M5). A participant proudly described her problem-solving traits: “If you want to know

how something works look at something someone else has already done and pull it apart

and put it back together yourself and that's pretty much ummm how I sort of approach

things” (M5). However, problem-solving was perceived as a trait that women might not

possess: “I guess a lot girls don’t go into programming at university. Umm perhaps they're

just not interested in problem solving” (G6). A stakeholder reflected:

Creative Problem solving

Transforming

Participation

I think they usually have an interest in tinkering with something, [...] It's the mechanics

of what makes it work that interests them and keeps them at their desk, problem

solving. So let's say if they weren't in multimedia or Flash they'd be in another problem

solving area... [interruption] it's kinda the ways brains wired. (S3)

Problem-solving manifested differently in different roles; for example, in

programmers finding coding solutions: “Programming, I can organise things, and follow

patterns and like doing it, it's like solving a puzzle” (G3). Problem-solving also manifested

in managing other people or clients’ needs; for some, it was “challenging when the client

says I want this and you try to make it something that meets their standards I guess

that's where the challenge comes from” (M6). Others enjoyed the challenge of bigger

picture problem-solving: “You know, technology's a very small part in a business, even a

technology business. Some guys just want to focus on that area. They'll get a problem or

a puzzle and they can keep their head down” (G8). One reason why the business side

was more appealing was that “it is more about people interaction” (G8).

Participants were motivated to pursue roles that offered them a challenge: “They

sort of spread us out to make sure we learn everything and can do everything, that’s kind

of what I’m interested in” (M8); and“I like challenging projects that make me think and

bring about new ideas, finding new ways to resolve them, improving myself over time.

There isn’t so much of that here” (M7). Similarly, another said: “I think it’s a positive as

well because that’s why I like my job... There’s something new, a challenge (M10). The

capacity for challenges influenced participants to pursue a career in the DCI: “It's

challenging, and it’s fun, it's creative. I couldn't imagine filing papers and answering

phones all day” (M6). Thus, occupational roles that did not offer participants challenges

were seen as unappealing:

People just do the same thing over and over again and then after awhile you can’t—I

don’t know. It doesn’t work your muscles in terms of your brain muscles, in terms of

your challenges, in terms of your problem solving. So I think that’s a big problem. (M7)

I get bored easily, what I won’t bored easily would be with computers and would be

because there’s such a variety with multimedia there’s always something different to do

in it, because you’ll always have to learn new things. That will keep me just hanging for

a long time. And so that’s why I went with multimedia. (M7)

Challenges motivated participants: “Whatever it is you do you always get to a

point where you're ready for any challenge” (G8); and“I remember it being really hard

because everyone else had used it before and I only just passed…But it was just the

challenge of it that drew me to it” (M6). Although they were often motivated by

challenges, the participants also valued some flexibility to manage these challenges; it

was “hard to work in a creative industry” (M4). One negative was the demand of learning

the new technology: “I guess one negative would be [inaudible] the learning curve again

and again and again” (M9). Others noted that “after being challenged so hard so long”

(M2), it could impact their health: “Sure it's great to be on the cutting edge - it’s more

creative & challenging - but it's also incredibly tiring and stressful” (G3). Certain roles

such as programming were more demanding, more “taxing because like the technology

changes every, I’d say 6 months to a year so you’re constantly reading and learning”.

However, she maintained that she “loved it” (G7). Another noted that when she was first

learning, it “was difficult, the workload was difficult but I thought that was fun” (M6).

Secondary sources also support the finding that the DCI is appealing to workers because

of the challenges and opportunities for innovation that it offers:

Game designers both women and men say the industry allows them to push technology in ways that they wouldn’t be able to incorporate IT departments. They also have growing opportunities to use gaming technology in innovative ways. (Appendix 17, Item 33) TRANSFORMING- The mechanism of transforming is difficult to describe as there

are few empirical examples of this type of creativity. When participants were asked to

provide an example of how they were creative, one described a creative act as a game

developer creating a character and bringing it life. Thus, it was interpreted that the

mechanism of transforming involves transforming entities such as artefacts, ideas and

processes. Participants chose DCI occupations believing they offered greater

opportunities for being creative than other industries: “I was interested in lots of different

things…They were photography, fashion design so creative type things” (G8). This

insight supports previous studies that report that students choose multimedia and games

because they are seen as more creative than IT (State of Victoria, 2007): “What I did

before I loved because it allowed me to travel but it was boring…The things that I didn’t

enjoy about it was that it’s not a creative industry, that was what I wanted” (G7).

Participants valued creative work.

There's so many different jobs in this industry, some are creative and some are not. I

think I wasn't fulfilling my creative needs quite early on. When I first got into it I was

saying I was very passionate about design and I was doing a lot of hours. I felt like I

was expressing a lot of creativity in this work. I think that disappeared fairly quickly,

having to design within the limitations of corporate. So it just became pretty much just

churn work, really. You tried as much as you could to put creativity into it but you're

quite limited. (G8)

Smaller “indie” organisations were seen as offering greater opportunity for

creativity. Larger organisations were seen as being “really risk adverse and so with the

games designs they just won't do anything innovative or exciting” (G3). Similarly,

government organisations constrained creativity: “I like innovation and I like finding new

things, doing new things, experimenting the new things, it’s really hard in a government

environment – they’re non-existent” (M7). Although one participant positioned creativity

in opposition to technical skill–“the course that I did at uni wasn't as creative…it was

very technical” (M6) –creativity could involve art or code; “to be good in any job you

have to be creative” (M1). Participants’ comments suggested that creativity manifested

differently across different interactive content creator roles.

5) HUMAN ‘ACT OF AGENCY’

The Human Act of Agency requires the person employ mechanisms such as

‘emotion’, ‘ethical’ and a sense of ‘being’.

Figure 18: The Human ‘Act of Agency’

EMOTION- The mechanism of emotion was evident when participants suggested

that emotions were an influence on their participation. A wide range of emotions were

identified by the female respondents, including: boredom, fun (G2), frustration (M4),

passion, fear, loneliness, desire, anxiety, anger, guilt (M1), and obsession (G7). For

example, one participant was ‘pissed off’ (interpreted as anger) that a male colleague was

earning more than her in essentially a similar role. Another described the ‘excitement’ of

working on a new project: “I’ve never worked on a realistic game...I was like this is so

exciting…” (G7).

The emotion most cited by participants was that of passion. Indeed, both

stakeholders and the extant data identified passion as a necessity to work in the industry:

“Like it's really hard to work in a creative industry so they have to be really passionate

and if they have that they're going to succeed, they will, they will succeed” (M4).

However, maintaining passion became a challenge: “It's a cycle in that you know

everyone is passionate about in work to put in these hours so then it's expected, and so

then it becomes the norm and then everyone else has to put in” (G6). Losing their

passion for the work led them to consider leaving the industry: “I just knew that I don't

have the energy and I don't have the passion for that any more. I certainly did but I grew

out of it, so I need to get out of it now before I am stuck here, having to learn the next

version” (G8).

Passion could counter negative influences on women’s participation, including

the lack of financial rewards the industry offered: “The money's not there, it’s really not

there...I think the only reason you would enter games because you've got a passion for

games” (G1). Passion was seen as integral to their lives and not only the workplace: “If

Emotion

Ethical Participation

you’re not following up a passion, I think that side of you will be unhappy and that

unhappiness will influence other bits, home life, relationship and everything like that”

(M4). Secondary sources support the importance of passion for DCI workers: “I believe

that talent, drive and passion [emphasis added] are the keys to success. Race, gender,

sexual orientation and the like are completely irrelevant” (Appendix 17, Item 28-, M, 32,

White, Ma, USA).

As noted earlier in the findings, there was a perception by both participants and a

stakeholder that men were more passionate about their careers than women. The level of

passion appears to be linked to the experience of playing games: “Passion also means

basically doing things around the periphery of that area. So even if it wasn't gaming but if

it was just doing volunteer websites or something like that in a non-gaming industry,

which is doing something that was external, I didn't find a lot of the girls were doing that”

(G8). The ‘guys’ were seen as being more passionate than women because they played

games and contributed to forums, "whereas the girls, no, never once did I see a really,

really passionate girl” (G8).

Developing an interest in playing games from an early age could support passion,

and it was suggested women had not been exposed to games in their childhood: “I guess

they just haven’t been exposed to it…it's a lot about how much you like games, games

you like, games you’ve played, how much you think about it” (G6). However, comments

from participants suggested that the belief that working in the industry entailed playing

games was a misconception: “I make music and sound effects for a range of different

products but I don’t play that many games” (S1); and“There’s always that mentality of oh

yeah, well I play games, I think I should go and make it games. It doesn’t work” (G7).

Other emotions included fun. Participants’ accounts suggested that simply having

fun supported their participation in the DCI: “I guess if it's something your friends are

doing then you’re probably more likely to get into it. It's fun because you share those

social experiences with other people” (G6). The Events of Interaction model drew

attention to other emotions such as anger regarding inequitable pay (the Catch 22 event

mentioned earlier), and the guilt of being a working parent (the Motherhood event

mentioned earlier).

ETHICAL - The mechanism of ethical was evident when participants suggested

that ethics were an influence on their participation. For example, when a participant

suggested bullying in the workplace as “not right” (M4), the researcher interpreted this as

an example of an ethical stance. Another example of ethics surfaced in a participant’s

account where she explained the value she placed on policies relevant to her role within

the government department: “Because it's government work we have to follow policy and

it makes me think I wonder if there's a policy on that...having that as a priority as well as

my work skills in what I do” (M4). The secondary data suggests that social value of

occupational roles is an influence within the ICT field:

Anyway, I got disenchanted with tech from a values standpoint… I didn’t think what I was doing mattered enough. So, I tried a career change. I went into Psychology […] and got a job in the field where I felt I’d be helping people. Turns out that field seemed even less ‘helpful’ to people than tech did. (Appendix 34) In the DCI context, participants made indirect reference to the social value of

their occupational roles. One participant, who worked as an online content editor (M8),

enjoyed her role because it facilitated a wide range of people being able to make

comments on a blog. Her background had been in journalism and she saw her current

role as facilitating aspects of participatory journalism. Similarly, a media producer

identified that a positive of the industry was that it sometimes creates products with

social value: “There are projects that are social projects like road safety and things like

that, it doesn’t always have to be advertising based” (M9). In contrast, one participant

who had been involved in producing reflected on the appropriateness of the violence in

certain games, and noted that her attitude to those games had changed since having her

first child: “I love violence in terms of the games that I play, I enjoy it…But I now have a

little boy… and for the first time ever I’m going oh, I don’t know about this. I mean it’s

fleeting ... So it’s like hmm” (G7).

Although certain aspects of the DCI, such as games content, are linked to what

might be seen as unethical aspects there is a grey area surrounding these. For example,

one stakeholder noted that the violence in games was being judged as being different to

the violence in movies. If a “movie depicted certain violence is okay but in a game

because – this is the difference – because you’re the one instigating that violent act, it

somehow falls into a different category” (S1). The hacker culture, which may be judged

as being unethical, can also be perceived as requiring talent. In reference to the hacking

culture a stakeholder stated: “I don't call that deviant I call that critical thinking” (S2).

BEING- The mechanism of being was evident when participants’ comments

suggested that other, less tangible, mechanisms were at play. The mechanism of being is

perhaps the most difficult to illustrate from the findings as it is a somewhat existentialist

concept. It is, however, best exemplified by the individual’s capacity for reflectivity.

Participants were reflective when they recounted their own experiences and reflected on

the thoughts and actions of others who influenced their participation. Further examples

are offered in Chapter 6, where the empirical data is linked to previous research and

theoretical work.

5.5 CHAPTER SUMMARY

The descriptive and explanatory findings presented in this chapter stem from the

logical, multi-level analysis of the empirical data. Application of each of the three

frameworks proposed in Chapter 3–1) analytical, 2) theoretical, and 3) ontological–to the

analysis of the data reveals different influences on women’s participation. In addressing

Research Question 1–What are the influences’ on women’s participation in the DCI?–the

analysis of empirical data revealed many influences; these included environmental

characteristics (such as the occupational stereotypes), the cultural practices and the

resources women had access to. The application of the frameworks has also helped

address Research Question 2–How might we understand these influences?–because each

framework not only allowed for the identification of different influences, but also

different interpretations of the influences on women’s participation.

As Section 5.2 revealed, application of Framework 1–the analytical framework–

led to the description of the characteristics of the Environment (E), Person (P), and

specific examples of Interaction (I) that may influence women’s participation. This

description has involved, as Danermark et al. (2002, p. 122) recommended, reporting

different empirical characteristics and types of events. Refining this initial framework led

to the development of the Sphere of Influence (SoI) model (as noted in Section 5.2.4).

The SoI model categorises the Environment (E) characteristics that influence women’s

participation in the DCI into four dimensions: social, cultural, mediated and resource

dimensions. Within these dimensions are properties which offer a further level of

granularity. Person (P) characteristics include a women’s social identity. Interactions (I)

between the E and P were evident in the events participants recounted, such as becoming

a mother. Ten events were encapsulated the Events of Interaction (EoI) model presented

on p. 163. Together the two models, the SoI and the EoI, began to offer a unified

approach towards describing the interactions between the environment and the person

that may influence a person’s participation.

As Section 5.3 revealed, application of Framework 2–the theoretical framework–

led to both the further identification of influences and a deeper understanding of

influences. Framework 2 was comprised of the Human Agency Multi-Theory Scaffold

(HAMTS) proposed in Chapter 3. Human Agency theories provided several concepts,

such as self-efficacy, for data analysis. These concepts assisted in the explanation of the

empirical data. For example, participants often referred to their self-confidence and how

this confidence could influence their participation. Bandura’s concept of self-efficacy

helped explain this confidence. There was empirical evidence to support all 11 of the

human agency concepts from the four agency theories in the HAMTS. Therefore, it is

evident that agency theories offer a useful perspective with which to investigate women’s

As Section 5.4 revealed, application of Framework 3–the ontological framework–

led to the identification of 10 agent-driven mechanisms that can influence women’s

participation in the DCI. A Critical Realist approach entails the identification of

mechanisms. There are ten agent driven mechanisms identified in the findings (see

section 5.4.1): 1) accessing, 2) imagining, 3) doing, 4) belonging, 5) sharing, 6) problem-

solving, 7) transforming, 8) emotion, 9) ethical and 10) being. These mechanisms are

organised into five categories of ‘acts of agency’, which are: 1) enabled, 2) connected, 3)

collaborative, 4) creative, and 5) human. These ‘acts of agency’, and the mechanisms

within them, are presented as a model of the Five Acts of Agency (FAA). These

mechanisms are driven by the person, in contrast to those that arise from the

environment. When women harness these mechanisms, it can result in actions that foster

their participation. The following chapter discusses how this identification was aided by

drawing on all three frameworks.

Although each individual framework provided a way to identify and understand

influences, the synthesis of the findings emerging from the analysis of data using the

three frameworks led to the original conceptualisation of agent-driven mechanisms. The

conceptualisation of these mechanisms is the key theoretical contribution resulting from

the research findings. The following chapter provides a greater level of theorising of

these findings with linkages between ‘concrete’ empirical data, extant literature, and an

emerging theoretical explanation.

Chapter 6. Discussion ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 211 of 401

Chapter 6: Discussion

Chapter 5 presented the findings resulting from the application of the three

frameworks (proposed in Chapter 3) to the analysis of the multiple sources of data

collected for the case study. The findings reveal that each of the three frameworks helped

in identifying and understanding the diverse range of influences on women’s

This chapter discusses those influences, within the context of the three

frameworks, and with reference to further extant research. The findings suggest that the

three frameworks can offer an approach to identifying influences. Many of the influences

identified using these three frameworks have been identified by previous research,

particularly research of women’s participation in the ICT context. Therefore, the

discussion of the influences draws heavily on this prior literature. There are several

nuances in the findings that are unique to the DCI context. For example, the outcome

expectations of women in the DCI include a sense of reward for a game product they

have developed attaining wide social approval or their own sense of pride in the work;

these rewards were rated by some participants as being much more important than salary.

Indeed, participants noted that DCI workers might even work for free just to have the

opportunity to help develop a game that they were proud of. This sense of pride as a

motivator has not been noted in the research stemming from the ICT context and in many

ways aligns to the pride an artisan may feel towards their craft. Furthermore, findings

indicate areas of research that require further investigation. For example the concept of

lifespan has been raised in previous research; the findings from this study further support

the value of considering temporality when investigating women’s participation. Perhaps

what is most notable about the findings is not the individual influences, such as the

media, but rather the extensiveness of the findings, which identify a wide range of

influences for the DCI context.

Importantly, this chapter also discusses the manner in which the three

frameworks offered a complementary approach to investigating the influences. Each

individual framework provided an approach to investigating the phenomenon of

women’s participation. However, the most salient contribution that emerges from the

research is the synthesis of the findings from each of the three frameworks, which results

in the conceptualisation of agent-driven mechanisms. Agent-driven mechanisms offer an

original approach to addressing the research problem: ‘Why do women participate in the

Australian DCI?’ This chapter contains a full discussion, interpretation and evaluation of

the findings; it shows how they advance our understanding of the research problem, and

their implications for the research area.

As Section 6.1.2 discusses, Framework 1–the analytical framework–identified

influences such as the environment characteristics, person characteristics, and events.

Environment characteristics include social contexts, culture, resources, and the media.

Person characteristics include social identity and, consequently, the stereotypes

associated with that identity. Findings also suggest that an influence may be the events

that a person experiences. These events may manifest serendipitously; for example, a

chance encounter with a mentor may provide an opportunity that is crucial to a person

gaining employment. As noted in Chapter 5, the refinement of Framework 1 culminated

in the proposal of two models that help reveal the influences on women’s participation:

1) the Sphere of Influence (SoI), and 2) the Events of Interaction (EoI). This chapter

discusses the value of these models to the research problem. As Section 6.1.2 establishes,

the dimensions and properties within the SoI offer a way to organise the diverse range of

influences identified from the empirical data. Existing research provides further support

for the conceptualisation of the dimensions, and strengthens the resultant model. Section

6.2.3 discusses the value of the Events of Interaction (EoI), a second model emerging

from the refinement of Framework 1. As the findings revealed, the EoI model identifies

10 events that may influence women’s participation. The discussion highlights that the

EoI provides a way to understand the (I) Interaction between the Environment and

Person. Both the models arising from the refinement of the analytical framework provide

a way to organise findings, and provide linkages to existing research.

As presented in Section 5.3 of Chapter 5, Framework 2–the theoretical

framework–was comprised of concepts from the Human Agency Multi-Theory Scaffold

(HAMTS). Framework 2 identified further influences arising from the individual

person’s characteristics, such as self-efficacy, and environment characteristics, such as

availability of role models, from the theoretical perspective of human agency. As Section

6.1.6 discusses, some of these influences have also been identified in previous studies;

however, agency theory has not been explicitly used as way to understand women’s

participation. The findings suggested that such a framework offers a useful approach. As

will be discussed in Section 6.1.6, the greatest value of employing the HAMTS was that

it assisted in the abstraction of mechanisms by drawing to attention the causal effect of

the individual agent. The reflection on the methodological value of employing the MTS

reveals its key benefits to the research process.

Framework 3 identified mechanisms. Mechanisms have not been identified in

previous related research as being an influence on women’s participation in the DCI.

As presented in Section 5.4 of Chapter 5, Framework 3–the ontological framework

(comprised of concepts from Bhaskar’s Critical Realism)–revealed five categories of

agent-driven mechanisms which manifest in the interaction between the person and

environment, and influence women’s participation. However, the identification or

‘abstraction’ of mechanisms has presented the researcher with a challenging task. As

Section 6.1.8 discusses, this challenge has been addressed in a number of ways, including

utilising both an abductive and a retroductive logic during data analysis.

As Section 6.3 discusses, individually, the three frameworks offer a

complementary approach to understanding the influences on women’s participation in

the DCI. It is, however, the synthesis of findings emerging from the three individual

frameworks that provides an original approach to understanding women’s participation.

A Critical Realist perspective, in conjunction with Human Agency theory and the

empirical data, leads to the identification of agent-driven mechanisms. The

conceptualisation of agent-driven mechanisms is an original contribution resulting

from the research presented in this dissertation. Agent-driven mechanisms offer an

approach to understanding women’s participation in the DCI.

As Chapter 5 showed, the empirical evidence supports the conceptualisation of

10 agent-driven mechanisms, as presented in the Five Acts of Agency (FAA) model (as

initially presented in Section 5.4.1 of the findings). Section 6.2.3 discusses the FAA

model, and the agent-driven mechanisms in a manner that addresses Stages 5 and 6 of

Danermark et al.’s (2002) model; this essentially entails linking the abstracted

mechanisms to empirical data, emerging theory and extant literature. Thus, further

empirical evidence is introduced in this chapter.

Section 6.3.3 presents further empirical evidence, emerging from this study, to

support the conceptualisation of the categories in the FAA model. This discussion further

reveals the agent-driven mechanisms that are important for women in the DCI context. It

might be assumed that this model may be generalizable to other contexts; however, this

would entail further research and application of the FAA model to those contexts.

Section 6.3.4 provides further synthesis of findings with the proposal of an

emerging theory entitled Acts of Agency. The theory provides an overall theoretical

outcome of the research, primarily by incorporating the previous models, including the

Sphere of Influence (SoI), Events of Interaction (EoI) and the Five Acts of Agency (FAA).

Where the FAA model has specific application to understanding women’s participation in

the DCI context, it is argued that the Acts of Agency theory is further generalisable to

other contexts of participation.

Section 6.4 concludes this chapter by discussing how the research goal for

applied research is addressed. This entails outlining how the agent-driven mechanisms

identified in the Five Acts of Agency may be used to propose and evaluate strategies that

foster women’s participation in the DCI context.

As Figure 19 illustrates, the two research questions are addressed in this chapter

through the discussion of the actual influences on women’s participation, and of how we

can understand these influences. The discussion illuminates a number of research

contributions, which include the development of frameworks that can help in the

identification of influences and in further understanding those influences. Further

contributions lie in the development of a methodological approach for the use of multiple

theories to scaffold inductive research that being the Multi Theory Scaffold. The

penultimate contribution lies in the use of the frameworks to develop the original

theoretical conceptualisation of agent-driven mechanisms and to present a model and an

initial theory to help address the research problem. Both the model and theory appear to

offer a valuable direction for further research.

Figure 19: Visual representation of this (‘Discussion’) chapter

Framework 2 (Theoretical/Human Agency theory): Theoretical Outcomes (Section 6.1.6-6.1.7)

Framework 3 (Ontological/Critical Realism): Explanatory Outcomes (Section 6.1.8-6.1.9)

Contribution: Influences include self-efficacy, norms and goals

Contribution: A Multi Theory Scaffold (MTS) offers several benefits; agency theories (the HAMTS) emphasise the role of the individual

Applied Outcomes (Section 6.3)

Contribution: The Five Acts of Agency model can be used to evaluate existing strategies and offer a basis for proposing strategies.

Contribution: Influences include underlying mechanisms

Contribution: Five Acts of Agency model identifies 10 underlying agent-driven mechanisms having a causal influence on women working in the DCI; synthesis of empirical, theoretical and ontological insights

Explanatory Outcomes (Section 6.2) Synthesis of resulting models into an emerging theory of Acts of Agency, which may be generable to other context and research domains

Framework 1 (Analytical/existing research): Descriptive Outcomes (Section 6.1.2-6.1.5)

Contribution: A range of influences are identified, including characteristics of the environment and person; several specific events are identified as being an influence

Contribution: Sphere of Influence (SoI) model organises influences; the Events of Interaction (EoI) model sensitises the researcher to the interaction between the environment and person

RQ2: How can we understand the influences on women’s participation?

RQ 1: What are the influences on women’s participation in the DCI?

6.1.1 WHAT ARE THE INFLUENCES?In answering the two research questions–‘What are the influences on women’s

participation, and ‘How might we understand these influences’–there is first a need to

clarify what is meant by the term ‘influence’. An influence is taken to be an entity that

emerges during the interaction between an individual and their environment, and has a

causal effect on an individual’s participation. Indeed, the term “influences” can be

considered “explicitly” causal (Gregor, 2006, p. 628). Findings suggest that an influence

can be a characteristic of the person or environment, an event, or an underlying

mechanism. A range of influences have been identified by using the three frameworks

for data analysis and several examples of these influences are presented in Figure 20. In-

depth discussion of these influences follows shortly.

It is vital that an influence be understood as a dynamic and complex concept.

Influences are not necessarily a barrier to participation, where the term barrier implies

finite closure; influences can either support or constrain participation. For example,

‘networking’, which has been noted in the findings and the extant literature, can be either

a supporting or a constraining influence. Networking can be a supporting influence when

it offers what Pratt (2006) describe as a pleasurable “hanging out” in creative industries.

Conversely, networking can be a constraint–a compulsory sociality (Gregg, 2008). Thus,

the influences identified in the findings are not definitive; they can manifest in different

ways in different contexts.

Indeed, one of the limitation of this study is that there is scope to clarify which of

these influences is unique to the DCI context. Many of the influences identified in Figure

20 have also been noted in previous research regarding the ICT context. For example,

stereotypes regarding women’s capacity for technology skills (as will be discussed in

Section 6.1.2) have been noted in the ICT related literature. This may in part be a result

of gender stereotypes permeating society beliefs. However, there are influences unique to

the DCI context regarding gender stereotypes. For example, many of the products that

are produced, such as games, portray women in a sexist manner. A future direction for

the research is to revisit the influences identified (for example, those presented in Figure

20) and to identify their specific manifestations within and relevance to the DCI context.

Framework 1

Framework 2

Framework 3

Environment

Social

o Family

o Peers

o Employers

Cultural

o Physical

o Ratio

o Language

o Economy

o Customs and

practices

o Legislative and

policy

o Historical

Mediated

o Websites

o Television

o Magazines

o Games

Resources

o Technology

o Information

o People

Individual

Social identity

o Gender

o Occupation

Events

meeting a mentor or contact

becoming a parent

networking opportunities

gaining skills

Giddens

Stereotypes

Clothing

Physical layout of

workplaces

Social norms

Connell

Skills and experiences as

sources of personal power

Bandura

Self-efficacy

Outcome expectation e.g.

rewards

Disposition

Self-efficacy

Vicarious experiences

Feedback

Mastery opportunities

Lent et al

Problem-solving

opportunities

Empirical

(as illustrated by Framework 1

and Framework 2)

Events

(as illustrated by Framework

Mechanisms (drawing on

framework 2)

10 agent driven mechanisms

1. Accessing

2. Imaging

3. Doing

4. Sharing

5. Belonging

6. Problem-solving

7. Transforming

8. Ethical

9. Emotion

10. Human

Figure 20: Examples of the types of influences on women’s participation in the DCI

In addition, an influence may be perceived or experienced differently by

individuals. As Lent et al. (1994) suggest, it is “likely that supports, opportunities and

barriers–like beauty–lie at least partly in the eye of the beholder” (p.106). Ideally, an

influence such as networking must be understood from the perspective of its meaning to

the worker (Gill & Pratt, 2008). What may be perceived as a barrier by one individual

may be a challenge, and even a motivator for another. For example, a participant

described how, when facing bullying in the workplace, the situation fuelled her to seek a

better career opportunity, whereas her colleague left the industry for the same reason.

Thus, two different individuals responded differently to the same influence of bullying.

An individual’s cognitive mechanisms and actions are integral to their participation as

they lead to the individual experiencing an influence differently. It has been suggested

that people with higher coping efficacy may perceive fewer barriers and greater supports

in their environments (Lent et al., 1994).

Although an ‘influence’ is dependent on both the causal capacity of the

environment and the individual, the individual’s causal contribution is central to their

participation. As Archer (2000) suggests, “conditional influences” may not in themselves

be “determinates of human action” (p. 131). A crucial point is that women are not to be

held entirely responsible for their participation. Women are not responsible for the

constraints, such as gender stereotypes, they encounter in their environment. Neither are

women passive recipients; rather, through their responses, they play an active role in the

subsequent effect of the environmental influences.

6.1.2 FRAMEWORK 1 (THE GUIDING ANALYTICAL FRAMEWORK): E, P, AND IAs Section 3.1 introduced, Framework 1 constitutes an analytical framework

developed from a review of previous research in the domain. The framework comprises

three categories: 1) Environment (E), 2) Person (P) and, 3) Interaction (I) between E and

P. As the findings revealed in Section 5.2, the refinement of the Environment (E)

category led to the conceptualisation of four dimensions (social, cultural, mediated, and

resource-related), and several properties (including language, practices and media) within

these dimensions. The Person (P) category included one dimension of the social identity

and two key properties (gender and occupation). These dimensions and properties

describe the characteristics of the Environment and Person that may influence women’s

participation in the DCI. In addition, the description of the Environment (E) and Person

(P) sensitises the researcher to the interactions between them. However, as the findings in

Section 5.2.3 revealed, the researcher reached an analytical impasse to further refining

the Interaction (I) category when using an inductive logic during data analysis. This

impasse was resolved by drawing on the ontological underpinning of Critical Realism

which directed the research approach to describe events. This lead to specific events of

interaction being identified.

The analytic process of employing Framework 1 culminated in the formation of

two conceptual models. As discussed in Section 6.2.2, the first model, the Sphere of

Influence (SoI) model, helps to describe the E and P influences. The second model,

entitled the Events of Interaction (EoI), helps to describe the interaction between the E

and P by illuminating 10 events that may influence women’s participation (as discussed

in Section 6.2.3). These models build on, and extend, previous research that has

considered the research problem of women’s participation. The development of these

two models is a major contribution of the research.

TEMPORALITY

The discussion of Framework 1, and the resulting models, would at first appear to

entail a consideration of the three categories (Environment, Person, and Interaction).

However, it emerged early in data analysis that the theme of ‘temporality’ arose across

all three categories. Participants often suggested that influences manifested in their past,

present, and future experiences. Accordingly, ‘temporality’ is discussed here prior to

subsequent discussion of the Environment, Person, and Interaction categories.

As the findings in Section 5.2.1 showed, both the female interactive content

creators and the stakeholders attributed historical characteristics of the environment as a

possible influence on women’s participation. For example, participants noted historical

events of an economic nature–such as the dot.com bust in 2001, and the GFC in 2009–

and attributed their current low salaries and limited employment opportunities to such

events. The influence of the dot.com event on new media workers has been noted in

previous research by Christopherson and van Jaarsveld (2005). Similarly, Trauth et al.

(2008) noted that in the US, fewer women than men returned to employment in the IT

field following the recovery of the “dot.com era” (p.9). Thus, history or historical events

may be an influence on women’s participation.

As Section 5.2.1 presented, women’s participation varies over their individual

life span stages. For example, participants suggested that gaining confidence with

computers at an early age has the causal effect of a greater self-efficacy in later years. As

occupational efficacy and job preferences are formed at a surprisingly early age

(Bandura, Barbaranelli, Caprara, & Pastorelli, 2001), inequitable access to computer

games may give males an unfair technological familiarity with both software and

hardware from a very young age in comparison to females (Huff, 2002, p. 112). Thus,

not having access to resources in early life stages may influence women’s participation in

later life stages.

Adopting a lifespan perspective not only identifies participants’ past experiences

but also their future decisions that may influence participation. For example, a majority

of the women in the study perceived the future event of having children as a possible

constraint on participation. Even though only two participants had children at the time of

the interviews, several others suggested they would need to leave the industry “in the

future when I have kids” (M7). A perceived incompatibility between future work and

parenting responsibilities may, as Bandura (1997) suggested, lead to participants

‘selecting’ themselves out of the industry.

A lifespan perspective has been identified in related domains, including career

development (Super, 1992; Lent, Brown, & Hackett, 1994) and, more recently, in the

ICT domain (Griffiths, Keogh, Moore, Richardson, & Tattersall, 2006). Agency

theorists, including Giddens (1993), also recognise “life course” (pp. 27, 82). The

research surrounding women working in the ICT industry has recently focused on a life

course perspective, recognising that metaphors such as the ‘leaky pipe’ may require

revision (Castano & Webster, 2011). A lifespan perspective can provide both a general

orientation for research, and add richness to insights by identifying causal relationships

and also foster holistic insights as it connects past, present and future influences.

The empirical data indicated that influences on women’s participation may also

emerge serendipitously through the emergent interaction of E and P within a situated but

dynamic context. For example, although one participant had an interest in computers

“most of her life” (G2), it was not until she met games workers in the coffee shop in

which she worked that she considered a career in the games industry. Had those

circumstances not eventuated, she indicated that she might not have entered the industry.

Agency theorists such as Bandura (1997) emphasise emergent aspects of interaction and

the role of serendipity. Similarly, even though Critical Realism recognises that history

may lead to certain structures predating actions (Archer, 1995), emergent interactions

remain central–even social structures and systems themselves are emergent entities

(Archer, 1995; Willmott, 1997; Wikgren, 2005). In part, this inclusion of a temporal

aspect addresses the need recognised in Chapter 2 for the relational analysis of

influences.

THE ENVIRONMENT (E) INFLUENCES

As the findings in Section 5.2 indicate, the Environment (E) category within

Framework 1 was refined during data analysis to include four dimensions: 1) social, 2)

cultural, 3) mediated, and 4) resource. Analysis revealed that within these four

dimensions are a number of properties such as social agents, language, and specific

media and resources. These dimensions and properties, as characteristics of the

environment, can influence women’s participation. This section discusses the four

dimensions in reference to the empirical findings and extant literature.

1) THE SOCIAL DIMENSION

The social dimension illuminates the social characteristics of the environment

that may influence women’s participation within the DCI. There are two key properties

in the social dimension: 1) social agents (including parents, siblings, teachers, peers,

employers and colleagues); and 2) social settings (including family, education, work, and

communities of practice). Findings indicate that social agents are those people that a

participant may interact with in a face-to-face situation (such as family, friends and

colleagues), and those with a greater distal proximity (such as role models). Findings

support previous research that identifies people who can influence women’s

participation; for example, fathers (Barker & Aspray, 2006), mentors in education

(Ahuja, 2002) and workplace contexts (Tapia & Kvasny, 2004), and teachers (Adya &

Kaiser, 2005) in the education context. In the wider literature, SCT recognises

socialisation agents such as mother, family, teachers, and peers are important role models

(Bandura, 2002b, p. 282).

Participants’ accounts revealed that certain social agents manifest in different

social settings or contexts; these social contexts can be positioned along a typical lifespan

perspective (See Figure 21). Social agents such as colleagues are more likely to manifest

in the social setting of the workplace, and parents (in particular, fathers) are more likely

to manifest in the family setting. Similarly, Giddens (1993) recognises that social agents

such as “family” in childhood, and peer groups in “later stages of an individual’s life” (p.

76) manifest in different “contexts of interaction” (pp. 78-79), including schools, the

mass media, and workplaces. Findings reveal that participation in a community of

practice, such as the online game World of Warcraft (WoW), could foster participation.

These CoP could manifest alongside other social contexts. For example, an individual

could be a student (education context) and also a member of an online gaming CoP. The

value of both formal and informal CoP has been recognised by Wenger, McDermott and

Synfer (2002). Further investigation of CoP is warranted.

There is scope for further refinement of the findings. Although findings did not

identify the family as being an influence in later life stages, it is likely the influence of

family may extend over a lifespan perspective, from the influence of parents in early

childhood through to the influence of children and spouses in late education and career

stages. The findings identified family as being most salient in childhood and early

education; this may be because children are more likely to be influenced by their parents

and that adults have a stronger sense of self and are consequently less likely to be

influenced by their spouse or child. However, family in all likely hood may be best

presented as a continuum over a lifespan. This an area for further research development.

Figure 21: Social settings and agents in the social dimension from a lifespan perspective

Social agents, such as family members, could be an influence on participation

because they could reduce the effect of social stereotypes. For example, Participant G3

(the AI programmer) attributed her participation to her parents not holding or teaching

her stereotypes about women’s capability. Conversely, parents could reinforce

stereotypes. For example, a participant remarked that women may not imagine

themselves using technology when parents reinforce stereotypes during childhood when

they “buy the son a computer and the girl a piano” (G2). These are similar findings to

previous research, which recognises that family members play a role in interpreting and

encoding external influences by reducing the salience of gender stereotypes (Oswald,

2008, p. 201). In the IT context, these stereotypes influence the career decision of girls

and women (Clayton, von Hellens, & Nielson, 2009). Thus, social agents such as parents

influence women’s participation in the DCI by conveying or challenging social

stereotypes pertaining to gender.

Social agents, such as role models, can manifest across different social settings.

However, findings also reveal that participants encounter a paucity of role models in

childhood, education, and career contexts. Women’s inequitable participation in the ICT

industry has been attributed to a similar paucity of role models in that industry

(McGrath-Cohoon & Aspray, 2006), and in education contexts (Thomas & Allen, 2006).

Role models are critical for girls (Newmarch, Taylor-Steele & Cumpston, 2000; Gürer &

Camp, 2002; Beekhuyzen & Clayton, 2004; Adya & Kaiser, 2005), and for women in the

IT industry (Tapia & Kvasny, 2004). Thus, findings identify similar influences to those

identified in previous research in the ICT and DCI contexts. However, as discussed in

Section 6.2.5, theorists such as Bandura offer further explanation of the influence of role

models.

2) THE CULTURAL DIMENSION

The cultural dimension illuminates cultural characteristics of the environment

that may influence women’s participation, such as language, customs and practices, and

historical conditions. Although participants made numerous references to how culture

could influence their participation, there were fewer insights to what culture entailed. As

Layder (1998) notes, there are differences between a participant’s colloquial use of the

term ‘culture’ and the manner in which the researcher may further interpret the term

when drawing on extant theory or concepts (p. 91). The researcher’s interpretation of the

term ‘culture’, and her understanding of how it can influence participation, is assisted by

drawing on extant literature.

Even though the term ‘culture’ is one of the “most widely” used concepts in

sociology (Giddens, 1993, p. 31), it is a difficult and complex concept to describe.

Indeed, culture is one of the most complicated words in the English language (Williams,

1983), with anthropologists advancing more than a hundred definitions (Berger, 1995). It

is often suggested that culture involves values, norms, and beliefs; however, there is

“wide-ranging and contradictory scholarly opinion” about these concepts (Straub, Loch,

Evaristo, Karahanna & Strite, 2002, p. 13). Hofstede (1993) proposed that culture

involves a shared cognitive framework; however, an exclusive focus on cognitive

components can have shortcomings (Sackmann, 1992). Culture has a multi-levelled

nature, including national culture (Trauth, Quesenberry & Huang, 2008), organisational

culture (Allaire & Firsirotu et al., 1984) and everyday situated contexts (Weisinger &

Trauth, 2003); it can also be inclusive of practices such as how members dress, patterns

of work, leisure pursuits, language (Giddens, 1997), and physical arrangement of the

environment (Kappos & Rivard, 2008). Although the concept of culture is complex,

there is value to the effort of trying to understand it (Schein, 1985).

The findings indicate that cultural influences manifest in different ways, at

different levels (macro and micro), and in different micro contexts. Culture differed

across organisation types; for example, government (public), private, start-up, or multi-

national. The definition of ‘private’ and ‘public’ sector organisations is taken from

Sethibe, Campbell and McDonald (2007) who also noted there has never been an easy

delineation between the two. There were few examples of macro level cultural

influences in the participants’ accounts. There was evidence of what can be described as

“subcultures” (Hofstede, 1998; Walsham, 1993). For example, a participant described

how the shared interest in pirates by a group of people in her organisation contributed to

a ‘nice culture’. A description of the ‘ratio’ property draws to attention the fact that DCI

organisations often had a male majority which, in turn, could lead to male orientated

practices–a “boys club” culture (S1, G3). Research in ICT-related literature suggests that

the ‘boys club’ can impede women’s participation in ICT (Margolis & Fisher, 2002;

Morgan, Quesenberry & Trauth, 2004). Similarly, research in the DCI by Roan and

Whitehouse (2007) suggests that an “all-male environment was not a problem”

[emphasis added], but it did offer a “different dynamic” and some “different issues” for

the most part, as male colleagues can “fear” that women “won’t fit into their [male

workers’] culture” (p.27).

Findings also show that cultural properties are evolving and dynamic. The

dynamic nature of culture is evident when we consider the practice of working long

hours. Although the participants worked on average a flexible, forty-hour week, several

women identified that there was an expectation within the industry to work long hours at

some point, and this practice could be an influence as it affected life/work balance.

Extant literature supports this, reporting a culture of long hours in the DCI (Gill & Pratt,

2008; Perrons, 2003b) and the ICT industry (Griffith & Moore, 2010). However, it is not

enough to suggest that ‘long hours’ may influence participation, but to consider how this

influence differs over time. Findings reveal that ‘cultural’ practices, such as working

long hours, can vary among different contexts, different times of participants’ lives, and

among participants.

There are different reasons why the practice of working long hours manifests.

Individuals may work long hours as a result of environmental characteristics, such as

economic conditions or management practices within the workplace, that lead to a

‘crunch time’ towards the end of project cycle. Although working long hours may be

driven by the individual (who is driven to produce excellence in their work for intrinsic

reward), this influence can vary at certain points of women’s careers. For example, when

they first start their careers or need to learn a new skill, they are more likely to invest

more time. Some participants made the point that they actively made the choice not to

work long hours in an effort to maintain a life/work balance, regardless of there being an

unwritten expectation that they would stay back like their colleagues. Thus, a cultural

influence, such as the practice of working long hours, is not a fixed entity: both the

environment and person have a causal role to play.

Perrons (2003b, 2005) suggested four main reasons why people work long hours:

the unpredictable nature and flow of work; uncertainty associated with a business start-

up; the need to continually update skills and knowledge; and the intrinsic satisfaction

derived from the work itself. Reeves (2001) proposed that working long hours may

reflect workers’ preferences; however, Perrons (2003b) criticizes the fact that the

demands of childcare are not considered in this suggestion by Reeves. Influences such as

maintaining a “work–life balance” have been identified as a negative influence on

women’s participation in the DCI (Perrons, 2003b) and ICT industry (Moore, Griffiths &

Richardson, 2005b). However, the practice of “work-life balance’ is changing, as

“quality of life has now become a term in the industry, and it was never a term before”

(Consalvo, 2008, p. 344) 79F

85 Culture is, as Schein (1985) suggested, a dynamic,

cumulative entity formed over time.

85 Game designer and Professor of Games Studies

3) THE MEDIATED DIMENSION

The mediated dimensionilluminates the influence of the media on women’s

participation. The analytic category does not suggest that media is influential in a direct

way, as might be suggested by an effects model. Rather, in response to criticisms of such

models (Gauntlett, 2001), this investigation adopts Silverstone’s (1999) approach that

media can involve ‘a process of mediation’, thus acknowledging the links between the

person and their environment.

The findings identify that the environment presents women with different forms

of media to interact with, including television and the internet. Previous ICT and DCI-

related studies have identified the influence of media on women’s participation. In the

DCI; these media include TV, advertisements (Gill, 2006), games content (Jenkins &

Cassell, 2008), and in the ICT context these media include magazines (Margolis &

Fisher, 2002). Similarly, agency theorists recognise the influence of electronic

technologies (Bandura, 2001; Bandura, 2002a; Bandura, 2002c), and television (Giddens,

1993). However, as Lang (2007b) laments, although the media is assumed to be a

powerful factor in women’s participation in the ICT industry, there is little data to

support the assumption, and its influence is under-investigated.

Mediums such as television perpetuate inaccurate gender and occupational

stereotypes of women. So much so, that it left one participant feeling “shocked”.

Secondary data supports the finding that gender stereotypes manifest in popular

television programs related to ICT and DCI. For example, The IT Crowd (a U.K.

Channel 4 television program) features technology savvy ‘geeky’ men and

technologically inept women. These insights are similar to those from previous research,

which suggested that “media images more frequently depict computer programmers and

developers as males and women as users” (Barker & Aspray, 2006, p.38). Mediums such

as television can offer women role models (Consalvo, 2008, p. 180), albeit fictional role

models. Appropriate role models on television have been linked to a “remarkable shift”

(that is, increase) in women’s participation rates in other industries, such as veterinary

science (Craig, 2004, p. 9). Participants in this investigation identified the paucity of

suitable female role models. Secondary sources of data indicate that sexist

advertisements in gaming magazines may contribute to the low level of female

readership which, this investigation proposes, may curtail women’s access to

information.

Nearly half of gamers are female. However, very few of them read gaming magazines. But what do you expect when you open the latest gaming magazine to a voluptuous blonde with a hardhat and a white cut-off halter-top...” (Appendix 17, Item 25)

Extant literature argues that women's involvement in the design process leads to

product diversity which then leads to attracting more women as both users and creators;

however it is not clear how. If there were more games that appealed to girls and women

this might lead to more women in the industry (Jenkins & Castells, 2008); this reflects

Fullerton et al.’s (2008) concept of the “virtuous cycle”. Secondary data supports this

notion that more women playing games may lead to more women in the industry. If

there's no games you like, you're not going to become a game developer” (Appendix 14,

Item 28 -M, 19, White, HS, Australia). However, Consalvo (2008) argued that there is

“no evidence” to support the notion that diversifying content encourages more women

into the industry (p. 177). Similarly, comments from the industry question whether

having more women in the industry would change the type of content in games.

Many games that are made specifically for girls involve role models that are insulting to parents and children. [Not that there are good role models in boys’ games.] Would this change if there were more women? Possibly, possibly not. (Appendix 17, Item 28-M, 24, White, disabled, Uni, Canada)

Previous research suggests other mediums, such as games products, may not

appeal to women because they perpetuate stereotypes or promote violence (Jansz &

Martis, 2003; Jansz, 2005; Ivory, 2006). Findings indicate that games content targets

men. A participant explains: “We know who our audience is and we do target them. We

target them by having hot chicks with big breasts and scantily clad” (G6). Participants did

suggest that games products may become less stereotypical with a greater number of

women entering the DCI workforce: “As more women do filter through to the industry”

(G6), and games are marketed to the “non-traditional gamer” (S1).86 However, participants

did not see such content as a barrier and happily admitted they played violent games.

Other influences– such as the positive social experience that accompanied the playing of

games or the opportunity to develop game play skills–fostered participation. This finding

highlights the complex interaction between the environment and the person influences;

and that more than one influence may be at play at any one time.

4) THE RESOURCE DIMENSION

The resource dimension illuminates the resources that may influence women’s

participation. As Section 5.2.1 introduced these resources include information,

technology and people. There is a distinction to be made between influences in the

mediating dimension and those in the resource dimension. A magazine may be a resource

when it provides industry information. However, when the content of the magazine

conveys stereotypes about a social identity, it then becomes a mediating influence. For

86 Examples of such products include Nintendo Brain Power

example, the gratuitous use of scantily clad females in industry magazines would

position the magazine in the mediated dimension.

Participants noted the importance of access to resources, such as computers and

magazines. Access to resources, such as computers, in early childhood fostered skill

development. For example, a participant (with access to a computer at home) taught

herself programming from a magazine at the age of eight. Access to resources was also

important later in life; for example, a participant described how she kept up to date while

on maternity leave because she had a computer and industry software at home. Many of

the resources the findings identify have been noted in previous related studies. For

example, computers have been identified as important resources for school-aged girls

(Beekhuyzen & Clayton, 2004) and women working in the ICT industry (Tapia &

Kvasny, 2004). In the DCI, Perrons (2003a) identified that new media employees engage

in self-learning via magazines and the internet.

The interrelatedness of influences is evident when we consider that information

resources could be facilitated by technology resources such as the internet. For example,

a participant identified that, although her school provided little relevant information

regarding DCI careers, she actively sought information from game development websites

such as gamedev.net, and gamasutra.com. This finding highlights that there is an

interaction between the environment and person influences.

THE PERSON (P) INFLUENCES

The Person (P) category in Framework 1 was refined during analysis to include

the dimension of social identity, which in turn included two key properties (gender and

occupation), and two further properties of ethnicity and age (as presented in Section

5.2.2). These characteristics of the person can influence women’s participation. This

section discusses the relevance of this finding in light of previous research regarding the

phenomenon.

1) THE SOCIAL IDENTITY DIMENSION

The social identity dimension illuminates the characteristics of the person that

can influence their participation. As Figure 22 illustrates, the Person category in the

original analytical framework was refined to include one dimension, that of social

identity. Within the social identity dimension are four properties: 1) gender, 2)

occupation, 3) age, and 4) ethnicity. It was interpreted that a person’s social identity is

the identity by which a person is most readily recognised and (possibly) judged.

Social identity is closely linked to stereotypes and social norms surrounding that

identity. As the findings suggest, participants’ accounts revealed they were aware of the

social norms and stereotypes surrounding their gendered and occupational identity. The

findings suggest that social identity may influence a person’s interaction with the

environment by fostering certain types of experiences, or by experiences. This

difference in experience occurs as a result of how other people interact with an

individual’s social identity, and how the person responds to that social identity.

Figure 22: Extending the person category

The term “social identity” can be further explained by drawing on Hogg and

Terry (2001) who described the social identity of a person as referring to the “social

category (for example nationality, political affiliation, organization, work group) within

which one falls, and to which one feels one belongs, and provides a definition of who

one is in terms of the defining characteristics of the category – a self-definition that is a

part of the self-concept” (p. 3). Thus, social identity appears important at the individual

level, manifesting as both self-concept, and a sense of belonging to a group. However, in

this investigation, social identity also involves the individual recognising the identity that

others see them as, and the stereotypes associated with that identity. Thus, the concept is

also similar to second order expectations which, as Webster and Whitmeyer (1999)

described, entail people responding to the perceived social expectations of others.

Although findings focus on social identity, identity is a complex and debatable

construct and may have several further elements or properties that these findings have

not revealed or explored in any depth. It may include aspects such as: national identity

(reflecting cultural characteristics); psychological identity (including traits and

personality); personal identity [which Trauth et al., (2004) propose is formed primarily

through a human’s own sense of meaning]; biological identity (age, health

characteristics); and organisational identity [which itself consists of personal,

interpersonal, and collective aspects (Brewer & Gardner, 1996)].

Indeed, a Critical Realist approach views the human as a “stratified being”

(Archer, 2000, p. 306), and recognises a number of levels of identity–such as biological,

personal, social, and cultural–as a constitutive element of a person’s identity. Critical

Realism requires that these levels remain analytically distinct to avoid the conflation that

occurs when a person’s capacities at one level, such as psychological capacities, are

Social Identity

Gender Age Occupation Ethnicity Properties

Dimension

Person

linked to biological capacities (Archer, 2000; Wikgren, 2005); for example, where a

woman's ‘personal’ capacity for leadership is founded on her ‘biological’ characteristics.

The focus on social identity, in contrast to other identities, may have resulted

from the methodological approach, which encouraged group level insights–in this case,

insights about women. Participants were asked about their personal experiences

(individual level), and about the influences on women’s (group level) participation. A

different methodological approach may have illuminated other aspects of identity.

Further investigation of the person category and the dimension of social identity is

warranted.

GENDER AND OCCUPATIONAL PROPERTIES: A DIFFICULT TENSION

As the findings reveal, the social identity of female DCI workers comprises two

key properties: 1) gender, and 2) occupation. Participants recognised their individual

capabilities were often judged against pervasive stereotypical notions of women’s

capabilities. Gender being one of the most pervasive ways by which we categorise

people (Stangor, Carr, & Kiang, 1998). Previous research reports that gender stereotypes

ascribe assumed masculine or feminine traits and capabilities, regardless of the person’s

individual qualities (Cejka & Eagly, 1999). Participants also recognised that the

occupations in the DCI were strongly associated with technical skills. Previous research

also suggests that the DCI values technical skills for development roles (Whitehouse &

Diamond, 2005). Findings reveal that women face gender stereotypes that suggest they

do not have the technical skills suitable for technology-related careers.

Participants faced the task of reconciling a difficult tension between the gender

and occupational properties. Previous studies reflect similar identity tensions: with young

women who were training to be IT teachers (Lloyd, 2009); with women in IT who face

dualisms such as ‘Home and Work’ (Beekhuyzen et al., 2003, p.77); and between

‘southern belle’ and the IT worker (Trauth et al., 2008b). Previous studies provide less

insight into how women respond to such tensions.

There are several entry points into further investigation of the tensions between

gender and occupational characteristics; these are: a) gendered skills (skills that are

attributed to genders); b) girly girls and geeky guys (gender stereotypes that suggest

women lack technical proficiency); c) passion and technical capacity (gender stereotypes

that suggest women are less passionate than men); d) the in-between gender (women’s

distancing from gender stereotypes); and e) gender triggers (the circumstances that

heighten the tension between gender and occupation).

a) Gendered skills Previous research reports that women face occupational segregation in the ICT

industry because certain skills are associated with a particular gender. Traits traditionally

associated with women in the ICT industry include emotional labour (Lewis & Simpson,

2007, p. 8) and communication skills (Lippa, 2002, p. 24); this has led to the

feminization of certain IT careers (Joshi & Kuhn, 2005; Cejka & Eagly, 1999). Even ICT

industry surveys, such as the 2010 ACS Women in the Industry Survey, ask questions

about participants’ use of soft skills (See Appendix 17, Item 41).

Findings suggest that women in the Australian DCI face occupational

segregation. As the findings reveal, women were associated most strongly with ‘soft

skills’ and administration and project management roles, and not with the technical skills

required for programming or production roles. Future research could investigate the

occupational segregation women face in the Australian DCI.

b) Girly girls and geeky guys Previous research suggests that technical proficiency in the DCI has been

traditionally seen as involving programming (Roan & Whitehouse, 2007). Even roles that

require artistic skills are seen as “a complement to underlying technical abilities rather

than hybridisation” (Roan & Whitehouse, 2007, p. 31). Findings reveal that these

technical skills are strongly associated with ‘geeks’, a term describing both an expertise

in an area and an unflattering persona (for example, unsociable, friendless). The findings

indicate that the most ‘geeky’ roles were technical roles involving programming and that

these roles were overwhelmingly associated with male workers. It appeared that the

gender stereotypes led to women, particularly those in technical roles, questioning their

capability with regard to technical skills. The women who were not technically skilled

were referred to by a participant as ‘girly girls’; this double emphasise on gender

highlighting that women were not associated with technical ability. Women in technical

roles (such as programming) faced the greatest challenges regarding stereotypes; this

may explain why there were few female programmers in the industry.

Previous studies also identify that it is very uncommon to have a female

programmer in the DCI (Roan & Whitehouse, 2007). Perhaps, as research regarding

women in the ICT industry suggests, women face influences that can cause them to

question their skill capability (von Hellens, Nielsen & Trauth, 2001, p. 118), and lower

their self-efficacy towards technology use (Betz & Hackett, 1997). Technology has been

inherently associated with “men’s work” (Trauth, 2002, p. 10), and the mastery of skills

is traditionally seen as a masculine trait (Turkle, 1995; Wajcman, 1991, 2000).

Similarly, secondary sources of data support the belief of women in the ICT

workforce that “a huge technology component [...] means: mostly men” (Appendix 17,

Item 31). Consequently, those women working in technology-related roles are somehow

“different” because they are going against society’s impression of what a female should

be (Turkle, 1995, p.56). In the IT sector, stereotypes contribute to an environment where

women are regarded as “other” to the norm (Lang, 2007a, p. 218).

Many of these beliefs are based on misconceptions, such as the notion that

women’s cognitive capacity for spatial reasoning limits their capacity in technical roles.

Hyde (1990) suggested that such beliefs are myths and any “gender differences in

cognitive abilities are generally not large” (p. 72).87 Future research could investigate the

stereotypes that female DCI workers, in particular programmers, face in regards to their

technical capabilities.

The findings from this investigation suggest that there is some ambiguity about

what constitutes ‘technical ability’. Participants often suggested that their roles were not

technical, when the term ‘technical’ surfaced in the interview; however, their roles

clearly involved technical skills related to technology. For example, the senior character

artist in a games organisation utilised software such as Photoshop and Zbrush, and a

social media coordinator required skills in online communities such as forums, blogs,

Twitter, and Facebook. The data indicates that even those women in roles more closely

aligned to administrative or managerial positions, drew on, and indeed took pride in,

their technical knowledge.

Women’s participation in the DCI may be constrained because the DCI does not

clearly communicate the diversity of roles available, and thus carries the burden of the

mis-informed stereotype that the industry only involves programming. Complicating this

matter is the rapidly changing nature of the skills required of the emerging industry. As

one stakeholder (who was a member of a peak industry body) noted, not even the

industry could define clear career roles or pathways. Future research could investigate

the stereotypes surrounding technical capability in DCI occupational roles.

c) Passion and technical capacity Being passionate, was an attribute a majority of participants identified as

important for DCI occupations. Indeed, both a participant and a stakeholder, who were

responsible for hiring workers, suggested an applicant’s passion was the deciding factor

in the hiring decision. Previous studies support this notion that passion is a key

87 There are some gender differences in one type of spatial ability, mental rotations, and mathematical performance. These were not statistically different. Hyde (1990) highlighted that trait ascription approaches have been challenged historically; for example, Wooley’s expose, in 1910, of the misconception that the smaller brain size of females negatively influences their cognitive ability. Wooley (1910) referred to the research of the time as “drivel and rot” (Hyde, 1990, p. 57).

characteristic of new media workers (Consalvo, 2008; Gill, 2002). However, the findings

from this investigation indicate that women were likely to be perceived as not being

passionate. Although participants were not able to clearly define ‘passion’, it appeared to

manifest as a result of early experiences with technology, such as playing games. There

was a suggestion that, as games content might not appeal to young girls, they would not

have had experiences with technology; this then led to their being perceived as not being

passionate about technology. However, as Section 6.3.3 discusses, the women in the

current study emphatically felt that they were passionate about their occupations and that

such passion could manifest in a number of ways. Future research could investigate

gender stereotypes that suggest women, in contrast to men, are not as ‘passionate’ about

technology. In doing so, there is a need to further investigate what ‘passion’ entails.

d) The in-between gender Previous research suggests that women in IT see themselves as “different” to

other women (Griffiths et al., 2005, p.12), where these other women are ‘typical’ women.

This distancing may be the result of male-orientated culture where women are forced to

fit into masculine cultures. In previous ICT research, it is suggested that “Female IT

workers are faced with two choices: either to masculinize themselves and ‘fit in’, or to

challenge the cultural system and attempt to feminize the workplace” (Tapia & Kvasny,

2004, p.87).

Participants’ accounts indicated that some women distanced themselves from the

stereotypical female identity. Participants described themselves as not being like other

women, they were, not ‘girly girls’, not interested in makeup, not “a particularly feminine

kind of person” (G7), or a “standard girl” (G6). This distancing from being ‘girly’ surfaces

in secondary sources, where, for example, a Girls’ Guide to Gaming explicitly states it is

not “girly” (Appendix 17, Item 8).

However, unlike previous research that suggests that women can only respond in

one of two ways to a male-orientated culture –by masculinising or feminising–findings

from this investigation indicate that women in the DCI take an ‘in-between’ stance. The

findings reveal that a majority of participants had a weaker gender identification, and

positioned themselves as neither a ‘girly girl’ nor ‘one of the boys’; rather, they were “in

between genders” (M2). The finding supports previous research that has suggested

women may be constructing a middle ground in regards to gender stereotypes. Data from

a 2002 study at Carnegie Mellon, for example, noted a “transitional culture” in which

some of the female participants “seemed to be constructing a new identity that was both

‘geeky’ and feminine” (Blum & Frieze, 2005, p. 112) –with ‘geeky’ being linked to the

masculine. Previous studies in the ICT field suggest fixed notions of gender are difficult

to support. Trauth (2002) drew on Wajcman (2001) who suggested that across different

cultures “there is no behaviour that is universally associated with masculinity or

femininity” (p. 101). Furthermore, traditional notions of gender stereotypes are possibly

changing with the younger generation (Trauth et al., 2010). Perhaps, as Bury (2011)

suggested, female geeks need to be understood as a hybridized alternative feminine

identity. Further research regarding women’s responses to the tensions between gender

and occupational identity may be warranted.

e) Gender triggers Although participants’ accounts indicated they could manage the gender

stereotypes they faced, tensions arose when the environment presented circumstances

where gender triggers were activated. Gender triggers are situational cues that prompt

divergent responses from men and women within specific situations (Bowles, Babcock,

McGinn, 2005). An example of such a gender trigger is evident in the inequitable male to

female ratio in the DCI organisations. Several participants noted that being in the

gendered minority in the workplace made them feel isolated and different. Similarly,

previous research in ICT suggests that being in the minority can leave women feeling

like the “odd girl out” (Trauth, 2002), and as being “treated differently” (Consalvo, 2008,

p. 188). Being in the minority can lead women to being evaluated first by their gender,

and then by their ability (Valian, 1998); and being the solo woman in a group heightens

gender stereotypes (Oswald, 2008, p. 201) which may have a potentially negative effect

on performance (Bandura ,1997).

Participants found that working in a male majority environment was manageable,

until an event or situation triggered the women into feeling as though they were treated

differently because of their gender. For example, participants felt like a ‘girl’ when a

male colleague restrained from swearing in front of a female DCI worker; this then made

her feel different and led to her self-doubt about her technical ability. Future research

could focus on the events and circumstances that trigger the tensions women experience

in regards to gender stereotypes.

AGE AND ETHNICITY PROPERTIES

In addition to the two key properties of gender and occupation, findings reveal

that age and ethnicity may also be, to a lesser extent, characteristics of social identity.

Participants perceived the DCI as being a young person’s industry; indeed, participants

ranged from 22-34 years of age. These findings support Gill’s (2002) observation that

one of the main attractions of the DCI is its image of “youth” (pp. 28, 78). DCI workers

can challenge traditional career progression pathways. Although age may mean seniority

in a traditional workplace, in the games industry, a forty-year-old may be retraining to

enter the newly emerging industry, while an eighteen-year-old already has several years’

experience in an upper management role or, as one participant suggested, as a “garage

CEO”.84F

88 A tension between age and occupational identity appears to manifest for older

workers who are in the minority within the workplace. One 30 year old participant

described herself as being “old” and of not “connecting” with her “20-something”

colleagues (G1). Influences can manifest as the DCI workforce ages. Life/work balance

may not have been an issue “at the age that they're looking at these roles” (that is, in their

early 20s); however, “in the older age groups” (S3), commitments such as parenting

surfaced. However, there is little research regarding the influence of age on the

participation of workers in the DCI.

Findings also reveal that ethnicity may be a characteristic of the social identity.

One of the two female programmers was from an Asian background; the other

highlighted that, of her female Computer Science classmates in a Canadian university, a

majority of the women were from an Asian background: “In computer science there was

definitely a higher percentage of male students; most of the females were exchange

students from Asia” (G3). Previous research identified a “striking predominance (around

40%) of Asian females among all female IT students in Australia” (von Hellens &

Nielson, 2001, p. 50). However, it is unclear if there is a similar pattern in the DCI. The

international IGDA survey (2005a, b) identified ethnic people, along with women and

people with disabilities, as minority groups within the games industry89.

It has been proposed that Asian women may pursue certain careers due to cultural

influences; specifically, to their Confucian family values and social expectations (von

Hellens & Nielson, 2001). However, as the findings (presented earlier on p.162)

highlighted that, although ethnicity may be a characteristic of the person, their

participation is not necessarily due to cultural influences alone. Rather, certain ethnic

groups entered the IT career pathway as a result of Australia’s migration policies. Thus,

it may well be a combination of influences such as cultural heritage and migration

policies that lead to the higher number of female Asian students participating in IT

studies in Australia. This finding illuminates the fact that influences on women’s

participation must be considered in a relational manner. Future research may consider

how the different properties of social identity (such as age, gender, occupation, and

ethnicity) can filter or influence participation experiences.

88 A garage CEO is a person who has given themselves the title of Chief Executive Officer of a company they operate from a home or a garage. The term can suggest illusions of grandeur but also has some credibility as several successful business people in IT started their own business in this manner. 89 The IGDA Diversity report identified a greater number of respondents self-identifying as ‘disabled’ than as ‘women’.

6.1.3 OUTCOME: SPHERE OF INFLUENCE (SOI) M ODEL

As the findings in Sections 5.2.1 and 5.2.2 reveal, the refinement of the E and P

categories in Framework 1 led to the identification of a range of properties, within four

dimensions of the environment (social, cultural, mediated, and resource-related), and one

dimension of the person (social identity). These analytical refinements presented in the

findings result in the development of a model entitled the Sphere of Influence (SoI), as

illustrated in Figure 23. 90

Figure 23: Refinement of the first two categories in Framework 1 resulted in the Sphere of Influence (SoI) model

There are five key strengths of the SoI to consider when evaluating its value for

understanding the phenomenon of women’s participation in the DCI. These strengths

include 1) the primacy of empirical data, 2) the linkages between macro and micro

levels, 3) a temporal perspective, and 4) a non-essentialist approach to gender, and 5) its

potential as a sensitising device for future research. Although the SoI does not

specifically include the analytical category ‘interaction’, the model does begin to

sensitise the researcher to the interactions between the E and P.

1) The first strength of the SoI is that it emerges from, and values, the empirical

data. The primary empirical data stems from the insights of the female interactive content

creators. Understanding the experiences of participants is congruent with Critical Realist

ontology, particularly with Bhaskar’s empirical domain of reality. Thus, the SoI differs

from models that emerged in later stages of analysis, including the Events of Interaction

and the Five Acts of Agency, which recognise influences that are not readily identifiable

by participants.

2) A second strength of the SoI is that it encourages a relational view of

influences. Even though the dimensions and properties in the SoI are presented as distinct

analytical categories, the findings reveal that there are linkages between the dimensions

90 The title of the model emerged from what Layder (1998) might call a “key interview “(p.71), where a participant described her male colleagues’ ‘sphere of influence’.

1. Environment (E) 2. Person (P)

of the environment (social, culture, mediated, and resource-related), and across both

micro (the immediate or situated setting, such as the organisation or workplace), and

macro (the wider social setting) contexts. However, further investigation of properties is

warranted.

3) A third strength of the SoI is that it encourages the viewing of influences from

a temporal perspective, by including an individual’s lifespan and historical events (as

Table 48 illustrates); this, in turn, fosters a unified and holistic understanding of

influences.

Table 48 Examples of the temporal aspect of influences on participation SoIDimension

Temporal Aspect

Characteristics Example of Influence

Cultural 1) Historical Historical events and circumstances which manifest in society and influence an individual’s participation

Affordability of technology (S2); women historically not in the workforce; recent events, for example, economic constraints, dot.com and Global Financial crises (G8)

Social 2) Lifespan An individual’s own lifespan, where personal events and experiences influence participation

Early childhood experiences with technology foster skill development (G3, M8); becoming a parent in later life can conflict with career

4) A fourth strength of the SoI is that the social identity dimension within the

Person (P) category provides a non-essentialist approach to investigating women. ‘Social

identity’ offers a concept midway between the individual and the stereotypes that women

as a group face (as illustrated in Figure 24).

- Women (essentialist)

- Group level characteristics

- Individual differences

- Methodological individualism

Figure 24: Positioning the analytical category of ‘women’ in the Sphere of Influence (SoI)

5) The fifth strength of the SoI model is that it can serve as what Giddens and

Turner (1988) described as a sensitising analytical device for future research. Used as a

sensitising device, the SoI is not a prescriptive model, but rather a tool to help identify

the characteristics of E and P that influence women’s participation in the DCI context.

The SoI model reduces some of the complexity in understanding participation, by

organising and unifying data. The SoI provides a holistic approach to understanding a

diverse and potentially fragmented range of influences, rather than what might be

described as ‘garden path analysis’. F

91 Hammersley and Atkinson (1983) suggested that

91 Richards (in Bazeley, 2007) describes ‘garden path analysis’ by stating that the reader is taken "along a pleasant pathway that leads nowhere: ‘Here are the roses, there are the jonquils, and aren’t the daffodils lovely today!'"

‘Social identity’ positioned between extreme approaches

such sensitising devices serve as a “germ of the emerging theory” (p. 180). As Turner

(Giddens & Turner, 1988) explained, until one has such a scheme “that denotes and

orders at the analytical level the properties of the universe”, it is “difficult to know what

to theorise about” (p. 163).

Thus, the SoI presents a starting point for early theorisation by addressing the

‘W’ questions of ‘What, When, Where and Who’. The ‘W’ questions ask: ‘What

influences exist’ (from the participants’ perspective); ‘When do the influences occur?’

(the historical and lifespan perspective); ‘Where do influences occur (for example, the

environment characteristics); and ‘Who is being influenced’ (for example, the person

characteristics). The importance of asking the ‘W’ questions has been noted across

different research approaches, including qualitative studies (Whetten, 1989), and case

study (Yin, 2003, p.5). Asking the ‘W’ questions during data analysis fosters a

“thoroughness of coding”, the development of “relational statements”, and a “theoretical

model” (Bazeley, 2007, p. 78). Answering the ‘W’ questions is useful in establishing the

range of any theory that emerges (Whetten, 1989), as it assists in identifying the

elements, concepts and relationships. For example, in asking ‘what’ influences exist, the

SoI provides categories, dimensions and properties to organise the descriptive insights

arising from the analysis of participant and secondary data. In asking ‘where’ influences

manifest, the SoI assists in recognising the characteristics of the settings in which

participation manifests–the “various conditions in which action occurs” (Grotz, 1999, p.

175), and the “settings of interaction” (Giddens, 1979, p. 207). The SoI does not

identify determinants of participation, but rather assists the researcher to recognise the

conditions, or the characteristics of the context88F in which participation occurs.

6.1.4 OUTCOME: EVENTS OF INTERACTION (EOI) MODEL

Developing a description of the Environment (E) and Person (P) sensitised the

researcher to their interaction. However, as noted in the findings, the researcher faced an

analytical challenge in further conceptualising and investigating interaction. This

challenge was resolved in part by identifying ‘events of interaction’ between the

environment and person, where these events appeared to influence women’s

participation. As Section 5.2.3 of the findings presented, the refinement of the Interaction

(I) category led to the identification of 10 events of interaction, which include gaining

entry into the industry (‘Getting In’), negotiating salaries (‘The Catch 22’), and

becoming a parent (‘Motherhood’).

The 10 events of interaction inform the development of the Events of Interaction

(EoI) model (See Table 39). As Figure 25 illustrates, the initial guiding categories of E,

P, and I offered by Framework 1 have been refined through the development of the two

models: 1) the Sphere of Influence (SoI) and 2) Events of Interaction (EoI). These two

models are contributions emerging from this research.

Figure 25: Development of the SoI and EoI models

There are four strengths of recognising events of interaction. First, identifying

events provides one way to understand the interaction of the environment and person–an

area that Chapter 2 recognised as needing a research focus. Second, recognising events

involves analysis of subjective accounts and objective insights (as inferred by the

researcher). This researcher inference can identify events not readily articulated by

participants. For example, a participant described an incident where she faced gender

stereotypes at work. While she described this as being a ‘small’ thing, the researcher

sensed that this incident had caused a strong emotional reaction (the participant was

visibly upset); thus the researcher interpreted the event as being an important one that

could potentially influence participation. This event was conceptualised as the ‘Muffins

and Men’ event. Pattern analysis confirmed similar events for other participants. Thus, a

third strength of identifying events of interaction is that linkages can be made between

individual level and group level analysis through cross level inference. This approach, of

identifying specific events of interaction, addresses in part the need for a multilevel and

relational analysis of influences, as identified in the literature review. A fourth strength

of identifying events of interaction is that it is an approach congruent with Critical

Realism, where the ‘actual’ domain within Bhaskar’s ‘three domains of reality’

recognises the events people may or may not be aware of.

As Table 49 indicates, several of the events identified in the Events of Interaction

model are similar to those in previous research. Indeed, a focus on events is not an

unusual approach, even from paradigmatic positions other than Critical Realism. For

example, from the interpretivist tradition Griffiths, Moore, Burns and Richardson (2007,

p. 34) recognised that particular events may “disrupt” the linear “flow’” of a woman’s

life history. A UK research project, ‘Disappearing women’, identified a range of

Events of Interaction (EoI)

3. Interaction (I)

1. Environment (E) 2. Person (P)

Sphere of Influence (SoI)

‘significant events’ that prompted women to leave the UK IT industry, including being

unable to return to the industry following redundancy (Griffiths & Moore, 2010). There

is scope to further identify similarities between the events identified in this investigation

and previous research.

Table 49 Linking the Events of Interaction to previous research Events of Interaction

Empirical example Linkages to previous research

Learning recipes Skill development Events such as gaining skills using the computer (Trauth et al., 2004)

Ladies’ lunches Social events to foster networking

Pub/after work drinks culture (Moore et al., 2005a, p. 18)

Are you the entertainment?

Male-orientated practices Games companies that throw recruiting parties with strippers (Braithwaite, in Cassell and Jenkins, 1998)

Mind your manners Gendered practices; for example, swearing

“...men that would swear and burp and pass wind and do all sorts of vulgar things...” (Trauth, 2002, p. 111)

Motherhood Women’s beliefs regarding the incompatibility of career and motherhood

Becoming a mother has been noted as an influence on women’s participation in the ICT industry (Quesenberry, Trauth & Morgan, 2006). In the ICT industry “Taking even relatively short career breaks” is seen as “risky” (Moore et al., 2005a, p. 12)

6.1.5 SUMMARY

This section discussed how Framework 1, the initial guiding analytical

framework (that emerged from the review of pertinent literature), has helped foster an

understanding of the influences on women’s participation. The key contributions

resulting from the analysis of the empirical data using Framework 1 led to the

conceptualisation of two models.

The first model, the Sphere of Influence (SoI), helps guide the description of the

influences by offering a way to organise empirical data into two categories of

environment and person characteristics. Environment characteristics manifest across four

dimensions: cultural, social, mediated, and resource-related; and across several properties

within these dimensions, including practices, technology, and social agents. Person

characteristics manifest as the one dimension of social identity, where properties such as

gender and occupation appear central. These characteristics may influence a women’s

Extending this description is a second model, the Events of Interaction (EoI). The

EoI model illuminates 10 events of interaction that may influence women’s participation,

such as becoming a parent. A focus on events can tell a richer and more complete story,

as it captures the dynamic nature of interaction between the environment and person.

Together, the models begin to offer a unified approach to identifying and

understanding influences on women’s participation.

6.1.6 FRAMEWORK 2: THE HUMAN AGENCYMULTI -THEORY SCAFFOLD (HAMTS) Chapter 2 identified the need for theoretical insight into the phenomenon of

women’s participation. In response, Chapter 3 introduced the value of using existing

theory in the form of a theoretical ‘scaffold’; and approach recommended by Walsham

(1995b) and by Layder (1998). Chapter 3 proposed the specific research approach of a

Multi-Theory Scaffold (MTS), a theoretical tool comprised of four types of theories:

meta, critical, middle-range and operationalised. The MTS was an early contribution

emerging from this investigation.

Framework 2 employed a MTS comprised of four existing Human Agency

theories (See Section 3.2); this is referred to as the Human Agency Multi-Theory Scaffold

(HAMTS). The four agency theories provided 11 initial concepts for data analysis (See

Table 41). The researcher’s reflection on the value of the HAMTS and MTS in this

section refines the contribution made earlier in Chapter 3. In answering the two research

questions–‘What are the influences?’, and ‘How might we understand them?’–there is a

need to evaluate the explanatory strengths of agency theory and to evaluate the

methodological implications of using a MTS in the research process.

The analytical process of employing Framework 2 during data analysis helped to

reveal a range of influences, as Section 5.3 presented in the findings chapter. These

included; stereotypes, clothing, physical layout of workplaces, social norms (using

Giddens meta theory); skills and experience as a source of personal power (using

Connells Critical theory); self-efficacy, outcome expectations e.g. rewards, dispositions,

self-reflectivity, vicarious experience, feedback, mastery opportunities (using Bandura’s

middle range theory); and goals, personal values and problem-solving opportunities

(using Lent et al’s operationalized theory).

Furthermore the analytical process of employing Framework 2 during data

analysis culminated in three further findings of a methodological nature. First, a MTS

provides the research process with certain strengths; for example, it can help sensitise the

researcher, offer concepts for analysis, and assist in empirically derived theorisation.

Second, employing a HAMTS is beneficial to the research process as Human Agency

theories offer a useful theoretical perspective with which to understand women’s

participation; providing concepts such as ‘self-efficacy’ helps explain the empirical data.

These first two findings were foreshadowed to some extent in Chapter 3. However, the

third finding–that agency theories are particularly useful in sensitising the researcher to

the underlying mechanisms that a Critical Realist ontology seeks to identify–emerged

from the data analysis presented in the findings and the theorisation that followed. The

first and second benefits are discussed in this section. The third finding regarding

identifying mechanisms is discussed in Section 6.3, as a part of the overall synthesis of

the three frameworks.

THE VALUE OF THE HAMTS

The HAMTS, proposed in Chapter 3, offered concepts from four different types

of agency theories with which to analyse the empirical data and understand the

influences on women’s participation in the DCI. There were several benefits of

employing the four agency theories in the HAMTS to the data analysis process. The

multiple theories fostered multi-level analysis; provided a wide range of concepts for the

exploratory research; provided an initial theoretical baseline of complementary

perspectives; and stimulated the theorisation process; leading to a deeper understanding

of the phenomenon of women’s participation.

The HAMTS revealed influences across multiple levels, such as societal

constraints, power relationships and individual cognitive mechanisms. The four theories

identified in Table 11 offer a more complete story than one theory. For example,

Giddens’ ST (a meta-theory) drew attention to social norms and power. Bandura’s SCT (a

middle-range theory) drew attention to influences manifesting at the individual level,

such as self-efficacy. As Giddens emphasises, both macro and micro level influences

must be recognised to make “connections” between them (Giddens, 1991, p. 112).

Recognising macro and micro levels may also be an important precursor to identifying

interactions (DiMaggio, 1991).

The HAMTS offered 11 concepts for data analysis; a wider range of possible

concepts than one theory might have offered. Furthermore, the researcher was sensitised

to a multitude of further concepts. Turner (1991) proposed that Giddens’ ST alone offers

11 sensitising concepts. Concepts that had not been used in the HAMTS became relevant

as analysis progressed. For example, the empirical data suggested that ‘problem solving’

was an important element of participants’ enjoyment of their DCI roles. The researcher

recollected that Lent et al., (1994) proposed problem solving as a possible mechanism,

not as a variable in their SCCT model, but as one of the 12 propositions put forth in

developing the model. Although the concept of problem solving was not one of the

initial concepts specified in the HAMTS, the early sensitisation to the concept was

beneficial as the researcher returned to the concept when it surfaced in the empirical data.

Several theories, in contrast to one, offer a wider range of initial concepts and this is

important for exploratory research that aims to avoid a closed approach.

The HAMTS offered a theoretical baseline from which to understand influences.

Drawing on agency theories during analysis assisted the researcher’s interpretation of

influences, particularly less readily visible influences, such as power. Previous ICT-

related research identified ‘power’ as being an influence on women’s participation

(Kvasny, Trauth, & Morgan, 2009). However, the findings show that participants made

few direct references to power in the current case study. The researcher interpreted

power by drawing on agency theory from the HAMTS. For example, a participant

described feeling excluded from work functions because they were held at strip clubs.

The meetings were seen as constraining the participant’s access to resources, and as

fostering the ‘old boy’s network’ which creates barriers to power for women. Giddens’

modality of Facility, and Connell’s STG, both suggest that exclusionary practices involve

power relationships. Similarly, participants identified a lack of mentors as being an

influence; this may also be linked to power as mentors can provide “reflected power”

(Kanter, 1977) and access to resources, such as information and contacts (Ragins &

Cotton, 1991). Thus, existing theory can serve as a “baseline” (Giddens & Turner, 1988,

p. 28), providing a point from which emerging conceptual or theoretical explanations can

be refuted or extended.

The complementary aspect of the four theories in the HAMTS fostered deeper

understanding of the phenomenon through triangulation. For example, participants

described feeling confident when they learnt the jargon associated with technical roles.

Concepts from the HAMTS help explain ‘confidence’. Bandura’s (1989b; 1997, 2001)

SCT explains that mastery skills may lead to a raised sense of self-efficacy. The

command of verbal skills is seen as contributing to power (Bryant & Jary, 1991, p. 10).

Connell’s (1987) STG draws to attention that jargon and skills may contribute

specifically to women’s participation by suggesting that “skilling and training is one of

the mechanisms by which the sexual division of labour is made a powerful system” (p.

100). Lent et al.’s (1994) SCCT provides examples of mechanisms that may be involved

in skill development. Thus, the different theories foster a deeper explanation of an

influence such as ’confidence’.

Upon reflection, employing the HAMTS has presented the researcher with

challenges; however, not to the extreme that Kincheloe (2001) described as the

theoretical “bricolage” leading to the “madness” of the researcher (p. 681). The most

significant challenge in using more than one theory is that it requires of the researcher a

great deal of knowledge with respect to several different theories; particularly in the

effort to juxtapose terminology and concepts in a meaningful manner. Certain theories in

the HAMTS offer particular challenges. Meta theories, for example, can be very abstract;

Giddens’ ST has been described as “quicksilver” (Craib, 1992, p. 11), a “theoretical

omelette” (p. 13), and having a conceptualisation of structures so “vaporous that it is next

to impossible to get a grip on them” (Healy, 1998, p. 510). Limitations in understanding

the theories can lead to a shallow explanation of influences. However, a HAMTS aims to

stimulate original insight by offering a toolbox for explanation; it does not offer the

finished product.

REFLECTION ON THE USE OF A MTS

The previous discussion considered the value of multiple agency theories during

analysis; that is, the value of the HAMTS. However, a MTS is not limited to agency

theory; other theories may be equally useful. As Chapter 3 noted, a MTS does not specify

which theoretical perspective should be used, but rather offers an approach to choosing

four types of theories to scaffold exploratory case study. When using a scaffold, “there is

no such thing as best theory only different ways of seeing the world” (Dobson, 1999, p.

261). Each of the four different types of theory: 1) meta, 2) critical 3) middle-range, and

4) operationalised, provides a different level of abstraction, and this can aid data analysis

and theorisation. This section provides a reflection on the value of using a MTS, in

general, in the research process. What is central in this discussion is that there is value in

employing more than one theory.

THE THREE BENEFITS OF A MULTI -THEORY SCAFFOLD

As Chapter 3 presented, not only is there “confusion” (Layder, 1998) regarding

the use of existing theory to understand empirical data, there is little or no guidance on

how to employ multiple theories as a scaffold. The findings indicate that multiple

theories may be useful because different types of theories, which offer different levels of

abstraction, lead to different insights. Previous research alludes to similar benefits. One

of the five types of theory Gregor (2006) proposed is ‘Theory for Explanation’; this

contains two subtypes of theories, including theory as a “sensitizing device” (p.624)–of

which Giddens’ ST is an exemplar. Gregor’s (2006) work not only proposed that

different theories offer different types of explanation, but that they may also offer certain

benefits to the research process, such as ‘sensitising’. In a similar vain, this investigation

proposes that a MTS may offer three different benefits to the research process. As Figure

26 illustrates, these benefits are that a MTS can: 1) sensitise the researcher, 2) offer

concepts for analysis, and 3) foster theorisation. Following is a discussion of each of the

three benefits.

Figure 26: Benefits of a Multi-Theory Scaffold (MTS)

1. Theories to sensitise(sensitise the researcher)

2. Theories to analyse (assist in analysis and interpretation of data)

3. Theories to theorise (assist in the theorisation

of findings)

1) SENSITISE

When we consider theories that sensitise the researcher, we are considering “prior

theorizing” and “empirical research” as “important inputs” (Miles & Huberman, 1994, p.

33). Theories that sensitise a researcher provide a theoretical orientation, rather than the

end point, for inductive research: ‘sensitizing concepts’ rather than a ‘ready to go’

explanatory framework (Patton, 2002). Employing the MTS sensitised the researcher to a

wide range of concepts that could help explain the phenomenon. Rowlands (2005)

suggested that such an initial sensitisation may occur in a preliminary literature review.

The findings from this investigation suggest that, although sensitisation occurs early on,

it may continue to be of benefit throughout the research process. During data collection,

analysis, and in later stages of theorisation, the different types of theories provided a

bank of concepts to draw on, as the empirical data suggested their relevance. Drawing on

a range of theories (rather than one) to sensitise the researcher was beneficial in the

exploratory case study as it initially encouraged a broad view.

2) ANALYSE

As Jones (1999) suggested, analysing with theory utilises a prior theory as an

analytical framework for the retrospective understanding of empirical situations or cases.

Theories identified in a literature review can provide provisional codes (Denzin &

Lincoln, 2008; Miles & Huberman, 1994). Researchers have used existing theory to

assist data collection; for example, where theory informs the structured interview

questions used (Walsham, 1995b). Existing theory can offer orientating concepts for data

analysis, where orientating concepts provide a “provisional means of ordering data”

(Layder, 1998, p. 108).These provisional codes can suggest a further line of inquiry, or

can be dispensed of in the “long run” (Layder, 1998, p. 110). These provisional codes are

open to revision, particularly when using theory for theory-building case research, where

the researcher is required to remain “open to alternative interpretations of data” to foster

“the development of new insights” (Anderson & Kragh, 2009, p.3).

In this investigation, analysis began from the provisional codes derived from the

three frameworks proposed in Chapter 3. The salience of these provisional codes was

refined as a result of the analysis of the empirical data. Application of the MTS assisted

data analysis in three ways.

1. The MTS provided linkages between empirical data and theoretical explanation 2. The MTS provided concepts that could stimulate the analytic process 3. The different theories in the MTS encouraged multilevel analysis

1. Employing the MTS provided the researcher with concepts that could link

empirical data to existing theory. As the findings indicate, each of the 11 concepts (from

the Human Agency theories in the MTS) could be matched, to some extent, to the

empirical data. This matching process offered a form of triangulation between theory and

the empirical data. Furthermore, those same theories could offer a step towards

explaining influences identified from the empirical data. For example, Bandura’s work

provides a body of literature with which to understand why role models may influence

participation. Thus, the MTS assisted in moving the research outcome from exploratory

and descriptive towards explanation.

2. Employing the MTS provided the researcher with concepts that could stimulate

analysis. As previously noted, the HAMTS drew attention to the influence of ‘power’.

Giddens’ ST emphasises that power is a causal capacity of society. For Connell’s STG, as

is the case with most critical theory, power manifests as a result of social structure and is

seen as belonging to the male majority, and being exercised to constrain women’s

participation However, participants’ accounts offered little direct evidence of the concept

of power. The empirical data challenged the theoretical concepts in the MTS, and led to

the researcher considering different explanations; that power may be expressed as a

personal power. Layder (1998) suggested that drawing on theory is useful when the

researcher runs ‘out of steam’. General theory, particularly Giddens’ ST, can introduce

“new lines of attack”, and new sources of creativity and inspiration, to the explanation

and interpretation of the data (Layder, 1998, pp. 124-125). Theories can assist by

“unfreezing” thinking through the juxtapositioning of the different perspectives or

“conflicting realities” (Eisenhardt, 1989, p. 546).

3. Employing the MTS assisted in the multilevel and relational analysis of the

data. Recognition of macro and micro influences is congruent with a Critical Realist

approach, which assumes mechanisms manifest across different strata such as society,

the individual, and biological and (even) meta-physical dimensions. Employing different

types of theory drew to attention a much wider range of influence, across more levels

than one theory could have.

3) THEORISE

To theorise involves the re-conceptualisation of data (Jones, 1999). This re-

conceptualisation requires the researcher “to scan, select and discard theoretical

perspectives” while “confronted with empirical data” (Layder, 1998, p. 3). Theorisation

involves thinking conceptually and analytically in order to answer ‘how’ and ‘why’

questions about a phenomenon, and providing explanation through specifying relations

and moving from particular to abstract–“to and fro between different ‘levels’ of reality

and analysis” (Layder, 1998, p. 100). In the conduct of the research in this investigation,

the MTS assisted theorisation in several ways, primarily the multiple theories

strengthened theoretical triangulation

1. Employing the MTS strengthened theoretical development. Employing several

theories, to be compared to each other and to empirical data, provides a form of

triangulation and this assists theoretical development, with a greater triangulation

offering more credible conclusions. In addition, employing multiple theories offers a

more developed explanation of the phenomenon. The MTS fosters statements regarding

the explanatory strength of existing theory, as it requires the researcher to consider the

limitations a theory has in regards to explaining a phenomenon. Theorising not only

entails a comparison of theory and data, but of theories themselves. As Lakatos (1978)

suggested, the process of refining theories must involve their evaluation in relation to

other theories (p. 13). Indeed, a Critical Realist approach involves the “immanent

critique of existing theories” (Yeung, 1997). Utilising existing theory may lead to the

modification of theory (Creswell, 1998), extension of a theory (Galliers, 1992), or

reformation of theory (Layder, 1998, p. 129).

The MTS provides several theories for comparison and, in doing so, can address

their respective limitations. Recognising limitations of the theories in a MTS can help

foster a deeper theorisation of the phenomenon. As Andersen & Kragh (2009) noted, by

using theory, it may become clearer how data and theorising add novel insights.For

example, Giddens’ ST (1984) recognises that a person’s knowledge and capabilities play

a part in interaction, there is a suggestion that ST’s substantive focus is on the “social

practices” (Rose, 1998), rather than on individual action and experience. If Giddens

illuminates the ‘aspects of the production of interaction’, it is proposed that Bandura

(1997) provides insights into the cognitive mechanisms, such as ‘self-efficacy’, that are

utilised by agents in that production. If Giddens’ ST suggests that agency provides a

space for agents to ‘act otherwise’, then Bandura’s SCT explains the mechanisms that

assist them to act.

Much of the work surrounding the development of theory from case study

recognises the value of using existing theory to strengthen insights. For Eisenhardt

(1989), tying emergent theory to existing literature enhances the “theoretical level of

theory building from case study research” (p. 545). In case study research, Yin (1994)

recommended not only utilising ‘a’ theory, but also rival theories. Walsham (1995b)

recommended the use of multiple theories to complement each other.

6.1.7 SUMMARY

This section discussed how Framework 2, the Human Agency Multi-Theory

Scaffold (HAMTS) helped to both reveal further influences and to offer theoretical

explanation of those influences. The HAMTS employed in this investigation was

comprised of theories relevant to Human Agency, including Giddens’ (1984) ST as a

meta theory, Connell’s (1987) STG as a critical theory, Bandura's (1986) SCT as a

middle range theory, and Lent et al.’s (1994) SCCT as an operationalised theory. The

discussion illuminated how the agency theories, in contrast to another theoretical

perspective, have been of particular value in developing a theoretical insight into the

phenomenon of women’s participation. However, there remains the possibility that other

theories may be of value in helping to understand the research problem. These other

theories may include other agency theories, or (even) theories not pertaining to agency.

This section also clarified the lessons learnt regarding the methodological

approach of utilising the Multi-Theory Scaffold (MTS) in general. The MTS was an early

contribution of the research, as proposed in Chapter 3. The discussion illuminated how a

MTS assisted the research process in three ways: 1) in sensitising, 2) in analysing, and 3)

in theorising. Although a single theory scaffold could offer a similar benefit to the

research process, the discussion emphasised the value of employing the multiple theory

approach. In a different research design, or with a different research problem, using more

than one theory may not be warranted.

A possible limitation of employing a theoretical scaffold is that the distinction

between existing theory and emerging conceptualisation may be difficult to make when

there is an intertwining between empirical data with apriori theoretical

conceptualisations. In response, this investigation aimed to clearly articulate the use of

existing theory. An account of how theory was used offers an open, reflective practice,

which strengthens the overall credibility of the findings and conclusions of the

investigation. Although the choice of which theories to utilise as a theoretical scaffold is,

as Walsham (2006) suggested, “essentially subjective” (p. 324), this section emphasised

that the choice of theories requires thought. In articulating the use of the MTS, there is a

greater clarity between the inductive process and the role of theory; this, in turn, assists

in the overall exploration, understanding and, ultimately, theorisation of the

phenomenon.

6.1.8 FRAMEWORK 3: ONTOLOGICAL (CRITICAL REALISM)Section 6.1.2 discussed how Framework 1 fostered a rich description of the

influences that manifest between the Environment (E) and Person (P), and their

Interaction (I). Section 6.1.6 discussed how Framework 2, a Human Agency Multi-

Theory Scaffold (HAMTS) moved findings from a description of influences towards an

explanation of those influences. This section discusses how Framework 3, the ontology

of Critical Realism, fostered further understanding of the influences on women’s

participation.

The discussion begins by reiterating what is required of a Critical Realist

explanation: essentially, the identification or abstraction of mechanisms. Although the

abstraction of mechanisms is a central aim for Critical Realist research, it also presents a

challenge to the researcher. The discussion describes how this challenge has been

addressed when abstracting the mechanisms presented earlier in the findings. The

mechanisms identified in Section 5.4.1 of the findings are not explicitly discussed in this

section. Rather, further discussion is reserved for Section 6.3.3, as the mechanisms

emerge from the synthesis of the findings arising from the application of the three

frameworks to the analysis of empirical data.

CRITICAL REALIST EXPLANATION

This investigation provides an explanation of the phenomenon of women’s

participation in the DCI. But what is meant by explanation? Explanation may involve

“explanans (universal laws, frameworks, conditions, triggering causes)”, and

“explanandum (description of what is to be explained)” (Danermark et al., 2002, p. 106).

A Critical Realist explanan involves explaining “why what happens actually happens”

(Danermark et al., 2002, p. 52); for example, when considering events (such as those

described in Section 5.2.3), the most fundamental aim is to identify a causal effect.

Critical Realism aims to answer the question, “What caused those events to happen?”

(Easton, 2009, p.4); this is because, as Sayer (1992) proposed, “causal powers are

important in any explanation”.

The findings from this investigation do not offer an explanation of the

phenomenon based on a classical causality of cause and effect but, rather, offer an

interpretation of causality from the Critical Realist ontological perspective. 90F

92 Such an

approach is pertinent to a study interested in interaction because, as Chapter 3 discussed,

interaction is dynamic. There is no “prescriptive” way to understand interaction due to its

unpredictability (Archer, 1998, p. 377). Unpredictability is in part a result of interactions

occurring across dynamic, open, stratified systems (Mingers, 2002, p. 299) in which

there are a multitude of mechanisms at play, contingent on other mechanisms, and which

may or may not be triggered. Sayer (1992) argued that mechanisms, being generative

rather than prescriptive, can sometimes produce “different events, in different

92 See Popper or Hume

conditions” (p.116); thus, “causality is dependent on conditions” (p. 106). We can,

however, say that certain objects “tend to act or behave in a certain way” (Danermark,

2002, p. 56). Therefore, an explanan involves probability, where “causal laws...must be

analysed as tendencies” (Bhaskar, 1978, p. 50), and analysis of patterns provides

empirical tendencies. Thus, the abstraction of mechanisms involves identifying what is

described by Lopez and Potter (2001) as the “tendencies of interaction” (p. 11). These

are not universal regularities, and a concept of fallibility is essential (Bhaskar, 1978).

THE CHALLENGE OF MECHANISMS

The Critical Realist emphasis on understanding causal tendencies requires the

researcher to identify mechanisms, to “establish the causal mechanisms that combine to

form the circumstances of the phenomenon observed” (Montano & Szmigin, 2004, p.

369). Unexplained phenomena can be explained by “proposing hypothetical mechanisms

that, if they existed, would generate or cause” the phenomena (Mingers, 2002, p. 300);

however, there are problems in the abstraction of causal or generative mechanisms

(Hodgson, 2004). The conceptualisation or abstraction of mechanisms is one of the

greatest challenges the researcher faced in the conduct of this investigation. There

are three reasons why abstracting or identifying mechanisms is challenging: 1) the

paucity of empirical exemplars of mechanisms, 2) the existence of mechanisms across

different levels, and 3) limitations surrounding the empirical observation of mechanisms.

1) The first challenge is that there are few Critical Realist empirical studies to

provide exemplars of mechanisms. Although Critical Realism emphasises that there are

enduring entities that have “powers or tendencies to act in certain ways” (Mingers, 2002,

p.299)–such as physical, social or conceptual entities– that to some extent, mechanisms

remain a “buzzword” (Gorski, 2008, p. 151), and “what is missing badly in the existing

realist practice ...is how such concrete research is actually conducted to examine

generative mechanisms” (Yeung, 1997, p. 70).

2) The second challenge is the multilevel and dynamic nature of mechanisms.

There may be a multitude of generative mechanisms at work, several mechanisms may

manifest at one level, and mechanisms may exist at lower and higher levels (Bhaskar,

1978). Mechanisms may also be relational and, therefore, research may even consider

interactions between mechanisms (Gorski, 2008, p. 159). Mechanisms must be

considered across different levels; this is because to overemphasise one level is to be

reductive, by employing a flattening ontological perspective (Wikgren, 2005, p.16).

However, Critical Realism “gives us little guidance to assess the importance of one

causal link compared with another” (Hodgson, 2004, p. 63).

3) The third challenge is that mechanisms may not be directly observable

(Gorski, 2008); that is, visible, perceived, or exercised. Although any “observability may

make us more confident about what we think exists”, that “existence itself is not

dependent on it” (Sayer, 2000, p. 12). Therefore, causal statements are made independent

of any empirical observations (Danermark et al., 2002, p. 55). Although mechanisms are

“reproduced continuously” they, or their effects, might only occasionally manifest

(Lawson, 1997, p.204). For example, mechanisms such as power may be unexercised or,

if exercised (at the level of the actual), may not be perceived in the level of the empirical

(Bhaskar, 1978).

Ideally, mechanisms in the ‘real’ domain can be identified if they have

manifested in the empirical domain (Tsoukas, 1989). However, because reality is an open

system, there are no regular conjunctions; that is, mechanisms cannot always be observed

or forced to manifest through experimental research processes, and there remains the

“impossibility of experimentation” (Yeung, 1997, p.53). The research approached aimed

to address these challenges.

APPROACHES TO IDENTIFYING MECHANISMS

In this investigation, mechanisms were abstracted in four keys ways: 1) by

reference to mechanisms within the Critical Realist literature, 2) by looking to the

broader literature for examples of possible mechanisms, 3) by drawing on Human

Agency theories, and 4) by employing retroductive logic during data analysis.

Underpinning this approach is Yeung’s (1997) recommendation that mechanisms be

abstracted through “careful theorisation and reflection” (p. 59).

1) There are few empirical exemplars of mechanisms in the literature from

researchers aligned to Critical Realism. However, two mechanisms are noted: goals and

reflexivity. Archer (2000) suggested that reflexivity is the “agent’s most important

generative mechanism” (p. 10). Danermark et al., (2002) proposed that goals are

important “human only” mechanisms, particularly when we consider that structures

cannot set goals (p. 179). Notably, reflexivity and goals are mechanisms that manifest

from the individual rather than the environment.

2) The broader literature also draws attention to a number of mechanisms,

although not explicitly from a Critical Realist perspective. Connelly (2001), in reviewing

an empirical study by Pawson and Tilley, proposed mechanisms that included: a lack of

incentive, inaccessibility and unavailability of jobs, and a lack of transport. Notably,

these mechanisms appear to manifest from the environment rather than the person.

3) Human Agency theories (in Framework 2) assisted the researcher in

abstracting the underlying mechanisms that Critical Realism (Framework 3) seeks to

understand. Agency theory is concerned with interaction between the environment and

person, and interaction implies cause and effect relationships. Agency theory makes

reference to mechanisms, although not specifically from a Critical Realist perspective.

These mechanisms included cognitive processes such as goal setting (Bandura, 1986),

self-efficacy (Lent et al., 1994), and social mechanisms such as power (Giddens, 1984;

Connell, 1987). Bandura described self-efficacy as one of the most central or pervasive

“mechanisms” of human agency (Bandura, 2002b, p. 270). Buchanan and Bryman (2009)

noted the relevance of Giddens’ ST for sensitising the researcher to generative

mechanisms such as trust, norms, beliefs, ontological security, and reflexivity.

As Figure 27 illustrates, the use of the theories in the MTS can support Stage 3 of

Danermark et al.’s (2002) Six Stage Model of Explanatory Research, which recommends

the use of extant theory (from an abductive logic) to assist in the abstraction of

mechanisms. This investigation reveals that the abstraction of mechanisms from

empirical data can be assisted by the HAMTS. Thus, a MTS is a useful addition to

Stage 3 of Danermark et al.’s (2002) model.

Figure 27: The MTS supports Stage 3 of Danermark et al.’s (2002) model

Figure 28 illustrates the concepts from the HAMTS that helped stimulate the

conceptualisation of mechanisms from the empirical data. For example, participants’

accounts emphasised that developing skills was an important influence on their

participation. Bandura’s concept of self-efficacy, and Lent et al.’s concept of self-

efficacy, also drew attention to the importance of skill development. The empirical and

agency theories helped the researcher abstract the mechanism of ‘doing’. The mechanism

of ‘doing’ involves the agent’s actions regarding skill development.

Stage 1 and 2 Stage 4, 5, 6

Stage 3

Extant theory

Inductive abductive retroductive

Figure 28: The HAMTS assists in the abstraction of mechanisms identified in the Five Acts of Agency

5) Mechanisms can also be abstracted from the analysis of empirical data by employing

retroductive logic. As noted in Chapter 3, Danermark et al. (2002) suggested that,

when utilising retroductive logic to identify mechanisms, the researcher may ask

questions such as ‘Could one imagine X without...?’ In light of the phenomenon

under study, the researcher posed the question: 'Can one image women’s

participation without the mechanism of x?’ This question is a form of theoretical test,

with the aim of identifying if an entity has a causal capacity and can, consequently,

be considered a mechanism. For example, the question then asked was: ‘Could

participation occur without women being ‘enabled’ through access to technology?’

As the empirical data suggested - access is a requirement- and accordingly it was

proposed that ‘access’ is a mechanism underpinning women’s participation.

6.1.9 SUMMARY

This section discussed the value of Framework 3 and, in particular, the Critical

Realist emphasis on identifying underlying mechanisms. However, this section also

highlighted the challenges regarding abstracting mechanisms. In response, the discussion

explained how the abstraction of mechanisms was achieved by drawing on existing

research–in particular, on Human Agency theories–and by employing retroductive logic

tests to the empirical data. In-depth discussion of the mechanisms identified in the

findings in Section 5.4 is reserved for the following section, as the mechanisms emerge

from the synthesis of the findings of the analysis of empirical data using the three

frameworks: the analytical framework (Framework 1), the agency theories in the MTS

(Framework 2) and the underpinning Critical Realist ontology (Framework 3).

6.2 A SYNTHESIS OF THE THREE FRAMEWORKS

This investigation proposed that the three frameworks outlined in Chapter 3

fostered an understanding of the influences on women’s participation in the DCI. The

previous sections separately discussed the value of each of the three frameworks. This

section illuminates how the synthesis of the findings from the three frameworks leads to

the specific focus on agent-driven mechanisms. Section 6.3.1 discusses the

complementary nature of the frameworks. Section 6.3.2 discusses how the three

frameworks fostered the conceptualisation of the agent-driven mechanisms. Section 6.3.3

details how the 10 agent-driven mechanisms, in the Five Acts of Agency (FAA) model,

can be used to understand women’s participation in the DCI. The discussion draws on

empirical evidence and extant theory.

6.2.1 THE COMPLEMENTARY VALUE OF THE THREE FRAMEWORKS

There are certain complementary aspects of the three frameworks that are

employed to analyse the empirical data; these, in part, stem from choices made in the

research approach adopted for this investigation. For the most part, each of the

frameworks provides an approach to revealing and understanding influences by fostering

an investigating into the interaction between the environment and person. As Section

6.2.1 explained, Framework 1 specifies the three categories of Person, Environment, and

Interaction. Although previous research focused on the first two categories, the review

identified the need to focus on the interaction between the two. As Section 6.2.5

explained, Framework 2, which is comprised of concepts from the HAMTS, also provides

an approach to investigating the interaction between the person and the environment.

Human Agency theories theorise interaction, as structuration (Giddens, 1979, 1984),

reciprocal relationships (Bandura, 1989b; 1997, 2001), and causal pathways (Lent et al.,

1994). As Section 6.2.6 explained, Framework 3, based on a Critical Realist ontology,

acknowledges the interactive nature of society and individual by recognising the

experiences of people, the events they experience, and underlying structures and

mechanisms. Thus, the complementary quality of the three frameworks strengthens the

overall research process and findings.

The complementary aspects of the three frameworks is further supported by their

relevance to Danermark et al.’s (2002) six-stage model of explanatory research which, as

Chapter 4 (Section 4.3.1) showed, guided data analysis. For example, Stages 1 and 2 of

the Danermark et al., (2002) model loosely aligns to Framework 1, as both foster

descriptive insights of entities involved in the phenomenon. Stage 3 aligns to Framework

2, as both employ an abductive logic when drawing on existing theory. The use of

existing theory is integral within a Critical Realist research process (Layder, 1998). Stage

4 aligns to Framework 3, as both use retroductive logic for the abstraction of

mechanisms. Therefore, both Stages 3 and 4 help reveal mechanisms. As Danermark et

al. (2002) noted, in practice, the “border between abduction and retroduction is not very

distinct as regards to concrete research” (p. 147). Application of Stages 5 and 6 of

Danermark et al.’s (2002) model results in theoretical explanation of the phenomenon.

Further synergies between the frameworks emerged during the conduct of the

research, and this strengthens the research conclusions. The first synergy was that the

frameworks supported each other during data analysis and theorisation. For example, the

researcher encountered a difficultly in conceptualising the ‘interaction’ category from

Framework 1. In response, the analytical object of interaction was explored inductively,

but with a focus on recognising Bhaskar’s events in the ‘actual’ domain, a concept in

Framework 3. A second synergy was that these frameworks foster both descriptive and

explanatory outcomes. Although Framework 1 helps describe the circumstances and

context that can influence women’s participation, a Critical Realist perspective

recognises limitations to descriptive insights and aims for explanation of a phenomenon

through the abstraction of the mechanisms at play. Recognition of the empirical

experiences (the empirical domain) and events (actual domain) is important; however, it

is a precursor to the identification of the underlying mechanisms. Bhaskar (1978)

suggested that “when a stratum of reality has been adequately described the next step

consists in the discovery of the mechanisms responsible for behaviour at that level” (p.

169). These mechanisms were identified by employing Frameworks 2 and 3. Thus,

Frameworks 2 and 3 move the findings from description to explanation.

Further synergies between the frameworks emerge when we consider the

similarities between the explanations that Human Agency theory (Framework 2) and

Critical Realism (Framework 3) offer. Critical Realism is compatible with agency

theories (Montano & Szmigin, 2004, p. 365; Dobson et al., 2007). There is a particular

parallel between Bhaskar and Giddens (Mingers, 2004), with Giddens’ work considered

to be “partly consistent” with Critical Realism (Danermark et al., 2002), and

ontologically compatible (Stone, 2001; 2005). Like Giddens’ Structuration Theory (ST),

the primary objects for Critical Realism are human beings and social structure (Lopez &

Potter, 2001, p. 15). Bhaskar (1979) emphasised that interaction is “the ‘point of contact’

between human agency and social structure” (pp.40-41). Giddens (1984) suggested that

interaction can be conceptualised as modalities linking structures and individuals.

Furthermore, both recognise the importance of the individual’s capability for action.

Bhaskar (1991) recognised the individual’s capacity to “reproduce or transform” aspects

of society (p.76), and Giddens (1984) emphasised the individual’s capacity to ‘act

otherwise’. Additionally, both recognise that individuals may or may not be aware of

their circumstances.

Furthermore, both Critical Realism and Giddens’ ST propose that paradigmatic

opposition between structure and agency are pointless. Giddens’ theory, in particular,

moves towards overcoming ontologically polar stances by bridging the gap “between

objectivism and subjectivism”, presenting any incommensurability of the paradigms not

as a dualism, but as a ‘duality’ of structure (Giddens, 1984). Such a duality of structure

allows for a multi-paradigmatic perspective (Weaver & Gioia, 1994) which, in turn, is

commensurate with the differentiated ontology of Critical Realism.

Although there are similarities, the stance taken in this investigation is that the

ontology of Critical Realism remains a central tenet in the use of Human Agency theory

and interpreting findings. As Archer (1995) proposed, ontological considerations may

“actively regulate the associated explanatory programmes” when considering agency

theories (p. 22). Therefore, points of contention are resolved by drawing on the Critical

Realist perspective in the first instance, in contrast to agency theory. However, such

points of contention may be embraced by the research process to make a stronger

theoretical argument. For example, when seeking to establish what is meant by the term

‘environment’, both Critical Realism and Human Agency theory offer insights.

The research took environment (a category in the SoI model) to be similar to

‘structure’, a term often appearing in the Human Agency theory and Critical Realism

literature. Structure is a fundamental concept because it highlights the importance of

setting in social processes (Reskin, 2003). Although a “basic and essential” concept,

‘structure’ is also a highly “ambiguous and contested” concept (Thompson, 1984, p.

156). The ontological status and explanatory power of ‘structures’ differs among the

methodological individualists, and methodological holists, realists and elitists (such as

Giddens) (Archer, 1995, p.105).

For agency theorists, such as Giddens, structures (rules and resources) exist only

in the instantiation created by the agent’s actions (acknowledged or not); they are

“virtual”, existing only in “memory traces” (Giddens, 1984, p. 17). For the Critical

Realist, Giddens’ conceptualisation of structures presents “ontological problems of two

kinds” (Archer, 1995, p. 107). A Critical Realist “stratified view of agency” would

consider both “prior structural conditioning and individual personality differences”

(Archer, 1995, p. 132). Structures must be seen as “real entities with powers, tendencies

and potentials” (Archer, 1995, p. 106); otherwise, they lack analytical significance

(Mutch, 2007). For Critical Realists, structures are not only instantiated by the agent, but

are also underlying organising principles, which may exist outside of social practices of

contemporary agents. Critical Realists do not see structures as ‘virtual’ but, rather, as

“actual forms of social organisations” (Dobson, 2001b, p. 206), such as “educational

systems, political parties” (Archer, 1995, p. 107).

Although Human Agency theory and Critical Realism differ in their

conceptualisation of structures, both agree that structures do not ultimately constrain or

predetermine the participation of the person. Structures can be thought of as a

“conditioning effect” that supply the “reasons for different course of action to those who

are differentially positioned” (Archer, 1995, p. 154). Although structures have a causal

effect, it is possible for individuals to either reproduce or transform the structures,

regardless of whether the person involved is aware of the structures. Recognising the

characteristics of the environment and the circumstances or structures that women face is

important; equally so, however, it is important to recognise how women overcome those

constraints, or (even) embrace supports.

The second point of contention is the concept of interaction. Although both

Human Agency theory and Critical Realism recognise the interaction between the person

and their environment, Critical Realism requires a clearer ontological distinction between

the interrelated categories of human action and social structure (Reed, 1997; Archer,

1995). Giddens’ conception of interaction as a duality is questioned by Critical Realist

theorists. Along with other ‘elist’ theories, Giddens’ ST is seen as being ontologically flat

(Archer, 1998; Emirbayer & Mische, 1998), conflating structure and agency and macro

and micro elements (Mouzelis, 1995). Giddens views structure and agency as mutually

constitutive (and hence inseparable) elements (Willmott, 1997), and is criticised for

blending the two (Layder, 1998). This conflation leaves little room for analytical

examination of the elements of the individual and society (for example, environment)

(Layder, 1998; Willmott, 1997), and limits the researcher’s ability to “examine their

interplay” (Archer, 1988, pp. 77, 80).

If Giddens suggested that there are two sides to a coin (the individual and

society), then Archer (1988) asked ‘How can one see both sides of the coin at the same

time?’ Archer (1995) suggested that the relationship between a person and society is

much more complex than theories that conflate ‘structure’ and ‘agency’ (such as

Giddens’ ST) can explain (p. 133). This may explain criticisms that “discussions of

agency to date have not dealt with different levels of analysis” (Hitlin & Elder, 2007, p.

Analytical dualism offers a methodological approach for Critical Realist research

(Archer; 1995; Dobson et al., 2007)91F

93, being the other “door” to examining agency and

structure, alongside ST (Willmott, 1997). 92F

94 Analytical dualism allows us to see “what is

particular about actors and particular about structures”, and also fosters examination of

interaction or the “interplay between” and the “interdependent, but emergent, strata of

social reality” (Reed, 1997, p. 31). This approach is preferable to making “the causal

powers of agents and their actualization impossible to analyse” (Fairclough, 2005).

Thus, Critical Realism can address the limitations levelled at agency theory

which, in turn, can strengthen emerging theorisation. In this investigation, the emphasis

placed on analytical dualism by Critical Realism led to a clear analytical distinction

being made when analysing the environment and person; more so than if solely using

agency theory. However, further consideration of how the concepts of environment and

interaction can be further conceptualised or theorised is warranted.

6.2.2 THE FOCUS ON AGENT-DRIVEN MECHANISMS

The previous section elucidated the complementary value of the three

frameworks. This section discusses the resultant focus on agent-driven mechanisms, a

concept resulting from the synthesis of the findings from the application on the three

frameworks; that is, a synthesis of the empirical data, previous research, Human Agency

theory, and the ontology of Critical Realism. Agent-driven mechanisms recognise the

causal effect of the person. As Figure 29 illustrates, the focus on agent-driven

mechanisms further extends the three categories of Environment (E), Person (P), and

Interaction (I) proposed in Framework 1.

93 ‘Dualism’ refers to the two strata of ‘social structures’ and ‘human agency’, and ‘analytical’ suggest these can only be identified by social analysis (i.e. not the everyday experience) (See Danermark et al., 2002, p. 181). 94 Archer’s own theory (1998) emphasises the need for analytical dualism.

Figure 29: Extending Framework 1 to include the agent-driven mechanisms

The focus on agent-driven mechanisms arose for several reasons, including

(simply) that the nature of the empirical data, which consisted of participants’ accounts

of their experiences, strongly alluded to the causal effect of the person. The women

interviewed emphasised how they had overcome the challenges they faced with regard to

their participation. However, the participants’ emphasis on the role the individual plays

in interactions was not captured in the findings resulting from Framework 1, which

focused more on the description of environment, person and events of interaction.

Framework 2, on the other hand, recognised the persons or agent’s capacity to act,

because agency theories inherently are concerned with how individuals might have a

causal contribution and challenge the contexts they are situated within. For example,

Bandura (1997) proposed that individuals can set goals, visualise their capacity, and

respond emotionally; this emphasised people’s own psychological functioning when

considering how they can “exercise influence” over what they do (pp. 2-3). Agency

theories propose that interaction involves not only the environment having an effect on

the person but, broadly speaking, the individual’s “generalised capacity to act” (Calhoun

et al., 2002, p. 85). Giddens (1984) proposed that human agents always “have the

possibility of doing otherwise” and to “act otherwise” in the face of structural constraints

(p. 258). An agent is an entity that can “act or exert power”, being a person “capable of

producing an effect: an active or efficient cause”.93F

Framework 3, Critical Realism, also helped recognise the role the individual

plays in the interaction between the environment and person, as it emphasises the

recognition of causal mechanisms. If Human Agency theory recognises a person’s

capacity to act, then the analysis of the empirical data from a Critical Realist perspective

encouraged identifying what mechanisms are involved in ‘acting’. If Human Agency

theory recognises the potentiality of the actions of agents, then Critical Realism helped to

identify the mechanisms the agent employs in actualising that potentiality. Giddens

(1979) emphasised rational actions and sought to understand agents’ “reasons for their

95 http://www.merriam-webster.com/dictionary/act

Frameworks 2 (human agency theory) and 3 (critical realism)

Environment (E) Person (P)

Interaction (I)

conduct” (p. 7). In contrast, the research findings align more closely to Bhaskar’s stance

that individuals may or may not be aware of the mechanisms involved.

The congruency between Human Agency theory and Critical Realism, and the

emerging emphasis from the empirical data on a person’s capacity for action, led to the

researcher reflectively asking the question: ‘If individuals have a causal effect, what

mechanisms are involved in their agency?’ In answering this question, the concept of

agent-driven mechanisms was developed, and the research focus narrowed to identifying

the agent-driven mechanisms involved in interaction.

Figure 30: Identifying agent-driven mechanisms by drawing on Critical Realism and agency theory

The identification of agent-driven mechanisms was further refined by asking a

second question when analysing the empirical data for evidence of mechanisms. The

question asked: ‘Does the action of the agent (as per their subjective experience and the

researcher’s objective insights) suggest a mechanism is at play?’ As Section 6.3.4

discusses, a focus on an agent’s actions helps to identify mechanisms that may not be

readily visible. Although an individual may not be aware of the mechanisms at play, it is

less likely that an individual cannot provide an account of their actions. Actions are more

likely to be observable by the researcher than mechanisms. Mechanisms, in specific

social mechanisms, will involve human activity (Gorski, 2008). Thus, the abstraction of

mechanisms involves recognising that the agent is doing something–an action, an act–

that has a causal effect. Accordingly, the conceptual labels of the agent-driven

mechanisms describe the possible action of the individual. The theoretical test asks

whether the agent is ‘doing, imaging, accessing, feeling, and collaborating’, and an

adverb or action word labels the mechanism.

Agent-driven mechanisms offer a perspective that contrasts previous research in

the ICT context, which has typically focused on barriers and the oppression of women.

Where previous research suggests that “circumstances...almost seem designed to wedge”

women “from the work they love” (Ramsey & McCorduck, 2005, p.7), agent-driven

mechanisms can help illuminate both the enabling and constraining influence of the

Critical Realism emphasises underlying mechanisms

Human Agency theory emphasises the person’s actions or acts

Society and individual

Everyday interactions

Action/Act

Experiences (Empirical domain)

Events and actions (Actual domain)

Mechanisms (Real domain)

individual. Previous research noted that women in the ICT field display “personal

optimism”, even when facing structural challenges that led to their feeling “pessimistic”

about the industry’s image (Moore et al., 1995a, p. 21). Recognising the causal

influence of the individual may help explain why women do participate in the DCI in the

face of adverse circumstances.

Agent-driven mechanisms can address, in part, the need recognised in the

literature review for research that acknowledges individual differences. The findings do

not suggest that every woman’s experience is underpinned by the same mechanisms.

Agent-driven mechanisms may manifest in different ways and different combinations,

dependant on the environment and the individual themselves. For example, a majority of

participants identified the need for role models. Role models can help women visualise

or ‘imagine’ their participation. However, this is not to say the mechanism of ‘imagining’

is at play for each individual. Other environment or person characteristics may lead to

women ‘imagining’ their participation; for example, women reading a book that

describes the industry, or an individual’s self-visualisation. The findings have identified a

tendency of certain agent-driven mechanisms within a particular group of women

working in the DCI. However, individuals themselves play a role in determining which

of the mechanisms they employ.

Agent-driven mechanisms recognise that women can, and do, challenge

constraints on their participation. Even though each individual woman interviewed

presented a different story of participation, the over-riding similarity of their accounts is

that they have overcome the constraints through their own actions. At times, the women

have been actively aware of how they fostered their participation (for example, actively

seeking information and mentors) and, at other times, serendipitous circumstances have

enabled their participation (for example, a chance encounter with someone in the

industry).

6.2.3 OUTCOME: THE FIVE ACTS OF AGENCY(FAA) MODEL

Agent-driven mechanisms recognise the causal effect of the person, in contrast to

those mechanisms that manifest from the environment or other agents. The abstraction

of 10 agent-driven mechanisms is a key, original, contribution emerging from this

investigation. Table 50 in the findings presented the 10 agent-driven mechanisms, which

are: 1) accessing, 2) imagining, 3) doing, 4) belonging, 5) sharing, 6) problem-solving, 7)

transforming, 8) emotion, 9) ethical, and 10) being. As illustrated in Figure 31 these 10

mechanisms manifest within five relational, non-hierarchical categories, encapsulated by

the proposed Five Acts of Agency (FAA) model. The FAA provides an original model

with which to understand why women participate in the DCI.

Section 5.4.1 in the findings offered empirical evidence supporting the

conceptualisation of agent-driven mechanisms. This section further illustrates how and

why each agent-driven mechanism influences the participation of female DCI workers.

In line with Stages 5 and 6 of Danermark et al.'s (2002) six stage model, the following

discussion of each of the mechanisms incorporates reference to theory (for the most part,

agency theory) and extant literature (for the most part, the literature related to women’s

participation in the DCI and ICT industry). As any discussion of findings from a Critical

Realist perspective must be done in reference to ‘concrete’ empirical data, further

empirical data is presented.

Enabled

1) ENABLED ACT OF AGENCY

The Enabled Act recognises that an

individual’s participation is fostered when they

access resources. There is one key agent-driven

mechanism, that of the agent accessing. Empirical

data indicated that participants accessed resources

such as technology (computers and software), information (job opportunities), and

people (mentors). Although there are three distinct types of resources, they are relational

as, in everyday life there, are no clear boundaries surrounding them. For example, to

have access to people resources might facilitate access to information resources.

ACCESSING: TECHNOLOGY

“When I grew up […] not everyone had computers” (M7).

The findings identify that, when women had access to resources such as

technology, this fostered their skill development. However, access to technology

resources differed over participants’ life spans. Access to computers in a participant’s

childhood was often a result of other social agents; for example, a brother bringing home

the family’s first computer. In the education life stage, participants noted that inadequate

access to technology resources within a school setting constrained skill development, and

consequently did not equip them with suitable skills for entry into university technology-

related programs: “You know, we had computer class and it was once a week and I think

it was only one year or something. Computers weren’t really integrated into the school

system” (M7).

In later life stages, participants’ access to technology improved, perhaps because

of their opportunities as adults working in the industry. Maintaining access to resources

whilst on maternity leave was a key strategy employed by women as it fostered ongoing

skill development. Participants’ access to technology may have also increased as a result

of technology becoming more readily available in the wider society over the last decade.

As noted previously, understanding influences entails recognising their historical context.

Agency theories offer support for the relevance of the agent-driven mechanism of

accessing. Findings reveal that access to technology can foster skill development which,

in turn, can help strengthen self-efficacy, a mechanism noted by Bandura (1997) and

Lent et al. (1994). Further extant literature identifies that access to technology is an

influence on women’s participation, including research that considers how and why

women adopt technology (Venkatesh & Morris, 2000; Ilie, Van Slyke, Green and Lou,

2005). Technological resources are typically noted in the ‘digital divide’ literature

(Castells, 2001; Kvasny & Trauth, 2002; Selwyn, 2002). Women IT professionals cite

access to computers while at school as the most prominent reason for their interest in an

IT career (Turner, Bernt & Pecora, 2002). However, Adya and Kaiser (2005) noted that

there are mixed messages regarding the influence of access to computers for school-aged

girls. It may well be that although technology is important, access has greater nuances

(Selywn, 2002), and that technology may be, as Bandura (2002a) suggested, “one

component embedded in an intricate network of sociostructural influences” (p.2).

ACCESSING: I NFORMATION

“It’s hard to really say definitively girls are being discouraged or whether there's just not enough information” (G2).

Participants identified the lack of information regarding roles, salaries and

opportunities (that fostered entry into the industry) as a constraint on their participation.

Access to information occurred through serendipitous or emergent events, such as

accidentally coming across a job opportunity on the internet. Alternatively, participants

employed active information-seeking behaviours; for example, by researching industry-

relevant websites. Previous research noted that industry websites such as gamasutra.com

are seen as valuable by DCI workers (Haukka & Brow, 2010).

The extant literature supports the finding that access to information (or the lack

of access) is an influence on women’s participation (Tapia & Kvasny, 2004); however, as

Archer (1995) suggested: “Knowledge alone is no spur to action (Archer, 1995, p. 131).

In post-industrial societies, ‘actionable knowledge’ is also a central resource (Cutler &

Company, 1996, p. 1). Hence, there is a link to other mechanisms such as doing, which

involves using the information.

ACCESSING: PEOPLE

“It’s the people you hang out with and your parents, what they move you towards or the options they open to you” (G6).

The findings indicated that access to people or networking could foster

participation. Networking is an action an individual takes to foster their participation.

However, male-orientated practices in the workplace could constrain such access. For

example, mentors were available to women working in the DCI; however, due to the

male majority, these mentors were most likely to male. Male-orientated practices, such as

conducting work-related meetings at the pub or strip clubs, were seen as constraining

women’s participation because they could exclude them from networking.

Agency theories support the relevance of the agent-driven mechanism of

accessing. Networking involves accessing people. Agency theories suggest that these

people, as “authoritative” resources (Giddens, 1984) or proxy agents (Bandura, 2001),

can provide access to other resources. Resources are not capabilities themselves; rather,

they facilitate capabilities, where these facilities help individuals “get things done”

(Turner, 1991, p. 524). There was for many of the participants an element of luck, or

what Bandura (2006) describes as “serendipity” or “fortuity” (p.166), in accessing

‘people’ resources; for example, in knowing someone (more often than not a male) who

had already gained employment in the industry (See findings in Table 44).

The extant literature identifies that access to other people is an influence on

women’s participation. Women working in IT report the “need for extra-curricular

socializing with male colleagues as necessary for career progression” (Trauth, 2002,

p.101). Women IT programmers noted negatively that too many “discussions took place

down the pub” (Moore et al., 2005a, p. 16), and that opportunities for networking occur

on “the golf course” (Trauth et al., 2008b, p. 19). Secondary data supports the finding

that the male majority can constrain women’s access to people and, consequently, career

opportunities.

My personal experience was that work lunches, afternoon golfing, etc. did not include women. That is one place where relationships are solidified, ideas are hatched, promotions are planned, etc. (Appendix 17, Item 31) Participants identified that one of the greatest challenges in entering the industry

was in knowing someone who could facilitate access, particularly as most employment

opportunities were ‘word of mouth’ (See Table 44 which summarises the method

participants used to gain access into the industry). The findings support previous

research, which suggests that access into new media employment is not the traditional

job advertisement in the paper, but informal networks (Gill, 2002). In the Australian DCI

context, an industry survey identified personal introduction as the most important way to

enter the industry (Haukka & Brow, 2010). Even entry as users of digital products, such

as games, involves access to social networks (Taylor, 2008, p. 53): “Most people come

into game culture through their networks and learn to be gamers within specific social

contexts”, usually through a “family member”, friend, or co-worker (Yee, in Cassell &

Jenkins, 1998).

Tapia’s (2006) case study of IT dot.com organisations found employees who

were hired through personal contacts (for example, friends of the owners) were the most

long-lived at each company. The fact that these networks can be a means of enhancing

confidence and offering moral support links the mechanism of ‘accessing’ to other

mechanisms, such as ‘connecting’, which involves developing confidence or, more

specifically, self-efficacy. Thus, the agent-driven mechanism of ‘accessing’ is an

important influence on women’s participation. Access can be fostered by the individual,

either through embracing serendipitous opportunity, or actively seeking out resources. As

Section 6.3.1 further discusses, the agent-driven mechanism of accessing can be fostered

by strategies that provide access to resources, information, and people.

Connected

Imagining Doing

2) CONNECTED ACT OF AGENCY

The Connected Act recognises that an

individual’s participation is fostered by their

identification with the DCI. The act of connecting

involves the two agent-driven mechanisms of

imagining and doing. These mechanisms are not so

much social connections to other people, but cognitive connections by the individual to

the industry and associated career pathways.

IMAGINING

“There isn't an image of a hot female programmer out there that people want to emulate or aspire to” (G10).

As the findings indicate, an individual visualising or imagining their participation

could be an influence. For example, seeing appropriate role models could prompt an

individual to imagine themselves ‘in the shoes’ of the role model–a role model being

someone that an individual could personally aspire to being (Holland & Lave, 2001).

However, findings also reveal that there was a lack of female role models in the DCI. In

addition, gaining insight into industry characteristics, such as salaries and lifestyle, could

influence participation because it helped the individual ‘imagine’ rewards. However, the

findings reveal that inaccurate information could lead to women not connecting with (or

being attracted to) the industry, as it made it harder for them to make informed decisions

regarding their career. For example, events such as becoming a parent appeared to be an

influence because women had no role models who were working mothers and often had

little accurate information regarding maternity leave. Consequently, several participants

could not visualise or ‘imagine’ themselves as being working parents; this led them to

surmise that they would need to leave the DCI when they had children.

Agency theories support the relevance of the agent-driven mechanism of

imagining. Bandura’s (1986) SCT drew attention to the importance of role models, and

suggested that vicarious observation of role models motivates an individual’s

expectations, goals, or aspirations. Such motives are the “wants that prompt action”

(Giddens, 1976, p. 85) in structuration theory. Lent et al.’s SCCT (1994) suggested feed-

forward mechanisms (as opposed to feedback) highlighted the importance of

anticipation, forethought and active construction of meaning (p. 87). SCCT also drew

attention to expectations such as materialistic (for example, monetary), social (for

example, approval), and self-evaluative (for example, self-satisfaction) outcomes.

Furthermore, agency theories other than those in the MTS support the finding that people

need to visualise what their future may hold. Emirbayer and Mische (1998) suggested

that, although the projective element is an important aspect of agency, the individual's

capacity for imagination “has been overlooked in the literature surrounding agency” (p.

971). Thus, agency theory concepts can help explain how women ‘imagine’ their

participation.

The extant literature also supports the notion that an individual’s capacity to

imagine or visualise can influence their participation. Previous literature identified that

negative perceptions of the IT industry, and a lack of female role models, can discourage

young girls from pursuing careers in IT (Lang, 2003; Lloyd, 2009; Beekhuyzen &

Clayton, 2004). An article reporting comments by the IGDA Women in Games SIG

panel laments “that young women could not imagine themselves as game designers”

(Appendix 17, Item 11). Thus, the agent-driven mechanism of imagining is an important

influence on women’s participation.

“It’s just by exploring and doing things” (M7).

As the findings indicate, gaining skills relevant to the DCI was integral to

women’s participation. Not only did gaining skills improve an individual’s ‘confidence’,

it could challenge the inaccurate stereotypes regarding women’s capacities. A sense of

confidence is central in pursuing one’s goals. Gaining ‘technical’ skills was pertinent to

women’s participation, as the DCI involves the use and development of technology.

However, further research regarding what ‘technical’ skills entail is required.

Agency theories can help explain why the agent-driven mechanism of doing

fosters women’s participation. Agency theories explain that the gaining of skills

influences, and can foster, women’s confidence to participate. Koesten, Miller, and

Hummert (2002) defined self-efficacy as “the belief that one has the capability to

mobilize the motivation, cognitive resources, and courses of action necessary to exercise

control over one’s environment” (p. 10). Thus, self-efficacy can drive participation. This

self-efficacy arises from experiences. As Bandura’s SCT explained, a sense of self-

efficacy or the self-belief in capability increases through mastery experiences; that is, the

individual practice and application of skills (Bandura, 1997).

The extant literature also supports the finding that an individual’s sense of

capability influences their participation. Previous studies in the ICT domain identified

that women may lack confidence regarding their technology-related skills, particularly in

late adolescence (Lippa, 2002, p. 29), and late education years (Pearl, Pollack, Riskin,

Thomas, Wolf & Wu, 1990). Self-efficacy is an important aspect of women’s capacity

when using technology (Betz & Hackett, 1997; Illie et al., 2005). Recent literature

regarding the ICT industry continues to recognise self-efficacy as an aspect of students’

participation and worker’s capacity (Kvasny et al., 2011; Joshi et al., 2010). Although an

Collaborative

Belonging Sharing

individual’s “natural ability” to work with computers can serve as a “mechanism to an IT

career” (Trauth et al., 2008b, p. 28), women also face fewer opportunities to gain

technological skills. Hence, many strategies aimed at improving participation rates focus

on skill development (Tapia & Kvasny, 2004). Skills relevant to the DCI may be learnt in

a variety of ways, where even gaming is considered a “gateway to mastery of a broader

range of digital tools” (Hayes, 2008, p. 218). Thus, there may be a link between girls

using digital products, such as games, and their participation in DCI career pathways.

Hayes (2008) added that when young girls gain mastery experiences in playing games, it

can lead to mastery of technical skills and “appropriation of tech-savvy identities” (p.

218). Taylor (2008) suggested that “demystifying play practises” may support self -

efficacy as it “increases women's and girls’ IT competencies and increases participants’

computer confidence” (p. 59). Thus, developing skills relevant to the DCI appears

integral to participation.

Even though gaining technology-related skills can foster women’s participation,

the ongoing demands of ‘doing’ technology can become a constraining influence.

Participants not only identified the need to develop skills, but also the need to continue

learning skills in the workplace context, as technology changes quickly. Being given

adequate time and resources to maintain those skills supported their participation.

Parberry, Kazemzadeh and Roden (2006) pointed out that games workers are expected to

learn independently, because the industry continues to push the boundaries of what can

be done using new computer technology. This may be because disruptive technologies

arrive more regularly in the DCI and are rapidly diffused (DCITA, 2005).

Thus, the agent-driven mechanism of doing is an important influence on

women’s participation. As Section 6.3.1 further discusses, the agent-driven mechanisms

in the connected act of agency category can be fostered by strategies that provide role

models, mentors and opportunities to develop skills relevant to the industry.

3) COLLABORATIVE ACT OF AGENCY

The Collaborative Act recognises the

importance of the individual having a sense of fit with

other people in the environment. The act of

collaborating involves the two mechanisms of

belonging and sharing.

BELONGING

“In order for us all do the best we need to work together and not work separate” (G3).

“If you feel like you don’t belong somewhere, you naturally start to question your reasons for sticking around”.94F

As the findings indicate, a sense of belonging or of fitting into the workplace

could influence women’s participation. Environmental characteristics that may influence

this fit include cultural norms and group dynamics within the workplace. Women often

felt they didn’t belong within the workplace due to being in the gendered minority; they

weren’t “one of the boys” (G2). Findings reveal that everyday cultural practices, such as

choice of clothing and open plan office spaces, could foster a sense of belonging.

Participants often faced challenges in fitting in because of their gender and the gender

stereotypes others held. Events such as “Muffins and Men” highlight that participants are

aware of the sanctions in the workplace that women face if they transgress social norms

regarding gender stereotypes.

Agency theories help explain why the agent-driven mechanism of belonging

influences women’s participation. Giddens’ ST (1984) drew attention to the agent

recognising the social norms that help them fit in or, conversely, the possible

repercussions of challenging such norms, including those linked to gender stereotypes.

As Giddens suggested, individuals reflexively monitor the risk of being sanctioned,

where this sanctioning can occur because of the prescriptions surrounding gender norms.

Further extant literature identified that a sense of belonging is an influence on

women’s participation. The DCI presents an organisational culture that emphasises that

workers fit into a team-orientated environment and work collaboratively.

Communication skills are valued in games production where there is a need to

communicate with people (Blake, 2011), and where close relationships exist with

publishers (Kerr, 2006, p. 75), fans, and users of products (Deuze et al., 2007). As Roan

and Whitehouse’s (2007) study revealed, employees “must be able to fit into the culture.

A low ego—there is no room for prima donnas here. A low ego, they must be able to

work with others. It’s all teamwork” (p. 30).

Secondary sources of data further support the importance of fitting into a team; if

you can “work well with the team, welcome aboard” (Appendix 14, Item 28-M, 33,

White, Uni, USA). Thus, as Castells (1997) suggested, being able to align one’s identity

with the organisational culture is a valuable asset.

96 http://www.lealea.net/blog/comments/women-in-tech-asking-the-wrong-questions/#ixzz0cP3g6Ng0 (Appendix 17 Item 31)

Previous research in the ICT domain suggests that it is important that women

gain acceptance in the organisation (Trauth, 1999, p. 159). However, an international

Women in Technology survey identified that 52% of participants stated that they worked

in uninviting workplaces where not being heard made them “question whether this is an

environment that they wish to stay in"(Appendix 17, Item 47). As Tapia and Kvasny

(2004) identified, “If people don’t feel important, they’re not motivated to stay” (p.88).

Environmental characteristics that may influence the individual’s sense of belonging

include organisational culture. Organisational culture can foster a sense of inclusion,

where inclusion is seen to involve ‘a sense of belonging’ when an employee is invited to

participate in important decisions, and feels that his or her input matters (Major, Davis,

Sanchez-Hucles, Germano & Mann, 2006). A sense of inclusion can lead to better

performance, greater job satisfaction, heightened commitment, and the increased

likelihood of remaining with the current employer and staying in the IT field (Major et

al., 2006).

The findings suggest that women could fit in, or feel as though they belonged, in

a number of ways; for example, by making an effort to socialise, to have lunch with

colleagues and to share interests, such as the unusual practice of being interested in

‘pirates’. Fitting in could also involve the individual learning the technical jargon

involved in DCI occupational roles. As Lave and Wenger (1991) identified, communities

of practice involve individuals learning to speak the vocabulary of the community.

A sense of belonging could also emerge from the clothing one wore in the

workplace. Due to people’s tendency to categorise each other (Maslow, 1962), outward

characteristics such as body and clothing can be an inescapable marker for the social

positioning of women. If the organisational culture fostered an environment where non-

gendered clothing (such as jeans and t-shirt, in contrast to stockings and skirts) could be

worn, this would provide a way for the women to not feel ‘different’ to the male

majority.

Thus, the agent-driven mechanism of belonging is an important influence on

SHARING

“You learn a lot through that team aspect” (M7). “I guess that's what I feel is lacking with my job, like having someone else to

talk to about that kind of stuff” (M6).

The findings indicate that sharing practices, experiences, knowledge, and skills

with others (such as colleagues) fostered an individual’s participation. In the workplace

context, sharing may involve individuals being mentored by others. However, the

findings suggest that being a mentor or providing assistance to others in the workplace

also provides a person with a way to share their experience; this, in turn, can foster their

participation. Participants suggested that being valued for their contribution to the team

fostered their participation, as it could provide intrinsic reward and supportive feedback.

People ask me questions and I ask them questions about stuff, which shares skills and

knowledge [...] I like to work with people and be able to bounce my ideas off people

and not worry about them ------ridiculed or anything. I just want to be able to do the

best job that I can and that includes being able to rely on the people around me and

also being able to know that I can point out something that they could do better umm

like the way their solving a problem or something. (G3)

Agency theories can help explain why the agent-driven mechanism of sharing

can influence women’s participation. Bandura’s SCT highlights that having access to a

mentor provides people with an opportunity for vicarious observation and scaffolding of

both skills and practices which, in turn, fosters self-efficacy and goals. Conversely, being

a mentor to others can provide intrinsic rewards or positive feedback which may

motivate and build the mentor’s confidence. Further extant literature also identifies the

influence of mentors in the ICT context (Tapia & Kvasny, 2004; Trauth et al., 2009).

The research findings from this dissertation suggest that sharing could be fostered

in a variety of ways in the workplace. For example, sharing could involve the sharing of

group level achievements. “You can see all of your work tying together, and you get a

real grip of how your work is affecting all the other work” (M10). And:

At other companies there is more of a collaborative environment I think we work less

hours but work stronger together. And umm a lot of companies I’ve worked at so far

have been like that [and is that important to you] Oh yes definitely I don’t like to work in

a completely super competitive environment where people are going to be a superstar

I'd rather work on something where we can be proud of it together. (G3)

Participants’ preference for open plan style offices can be explained by noting

that this type of physical structure encourages transparency, collaboration, and sharing of

practices as well as space. Fullerton et al., (2008) suggested that the ideal work

environment for women is one with open space. DCI organisations are seen as providing

“relaxed and non-hierarchical” working “environments and relationships” (Gill, 2002,

p.78). Thus, the agent-driven mechanism of sharing is an important influence on

in the collaborative act of agency category can be fostered by strategies that foster non-

discriminatory cultural work practices, and those that foster communication between

workers.

Creative

Problem-solving Transforming

4) CREATIVE ACT OF AGENCY

The Creative Act recognises the importance

of the individual having an opportunity for

creativity in their occupational role. The act of

being creative involves the mechanisms of problem

solving and transforming. Creativity is a highly

valued characteristic in the DCI context, both by employers as it is seen as fostering

innovative practice, and by workers who often obtain an intrinsic reward from their

creativity. Almost every participant identified that a positive aspect of their DCI careers

is that it could provide them the opportunity to be creative.

PROBLEM SOLVING

“You get people that are driven to find out how things work and I want to do that myself you know” (G6).

As the findings indicate, participants found an occupational role appealing if it

required problem-solving skills. Problem-solving provided challenges which, when

resolved, offered intrinsic and extrinsic reward. Being a problem-solver was seen as an

integral trait of a DCI worker. The problem solving appeared to be specifically tied to the

workplace, and often surfaced as a result of the rapidly changing nature of technology.

However, participants also recognised that responding to these challenges (such as

learning new skills and working on cutting edge projects) required a time investment that

could distract from a life/work balance. The findings support similar insights in the ICT

industry, where Gallivan (2004) identified that IT professionals face a stress associated

with the rapid technological changes characteristic of the industry. However, problem

solving was not always linked to technology; it could involve participants meeting a

client brief in an innovative way or developing a creative design. Different types of

problem-solving manifested in different roles.

Agency theories can help explain why the agent-driven mechanism of problem-

solving can influence women’s participation in the DCI. Bandura’s (1989b; 1997, 2001)

SCT proposes that intrinsic rewards are a result of a person achieving their goals, and

solving problems provides such a reward. SCCT proposes that problem solving is a

strategy involved in skill mastery (Lent et al., 1994). Smith (2002), who drew on SCCT

to explain career participation in ICT, suggested that people in IT are “individuals who

use the computer as a problem-solving tool of choice” (p. 1). Previous research noted

that problem solving can be an influence on an individual’s choice to participate in ICT-

related careers. Agosto, Gasson and Atwood (2008) stated that “most studies of why

women avoid IT have demonstrated that females who are making career choices care a

lot about ‘solving problems’ and much less about ‘technology’” (p. 205). The

importance of women in IT having “challenging workplace projects” has also been

recognised (Trauth, 2002, p. 109). Tapia and Kvasny (2004) have recognised that IT

employees want to learn, sharpen their skills and hone new ones. The situation is similar

in the DCI. A female games industry veteran of 17 years noted that creativity and

problem solving are key skills (Blake, 2011)

Studies of IT professionals identified that women rated ‘wanting the opportunity

to learn new things’ more highly than they rated ‘salary’. Wanting to ‘gain more

expertise’ and not face ‘boredom’ rated more highly than being ‘discriminated against’

or ‘not fitting in’ (Hoonakker, Carayon, Schoepke, 2006). Interestingly, recent research

suggested that the capacity for problem-solving may be seen by younger people (college

students) as being a gender neutral trait (Joshi et al., 2010). The findings support this

previous research; several women expressed the lack of challenge as a constraint on their

participation.

Thus, the agent-driven mechanism of problem-solving is an important influence

on women’s participation.

TRANSFORMING

“It’s not a creative industry and that was what I wanted” (G7). “To be good in any job you have to be creative” (M1)

Although there are challenges in establishing a definition of creativity from the

extant literature, it is generally agreed that creativity involves the generation of ideas that

are novel and of value (Gruys, Munshi, Dewett, 2011). Creativity involves the

individual’s actions (Joas, 1997; Emirbayer & Mische, 1998). Although individuals are

“the inventors of new possibilities for thought and action” (Emirbayer & Mische, 1998;

Joas, 1993), they often draw on pre-existing resources. “Creativeness rarely springs

entirely from individual inventiveness”, but rather “by refining pre-existing innovations,

synthesizing them into new ways and adding novel elements to them something new is

created”(Bandura, 1999b, p.33). Therefore, it is proposed that the term ‘transforming’

reflects that creativity often involves transforming something.

As the findings indicate, a majority of participants spoke of the appeal of the DCI

being a creative industry. Being creative included creating visual designs for web sites,

and creating characters and storylines for games. However, there were few empirical

examples of how this creativity manifested in everyday workplace situations. It appeared

that participants may well enter the industry with the hope of being creative, but find that

the certain organisations and work practices constrain opportunities for creativity.

The findings also suggest that opportunities for creativity varied with roles and

types of DCI organisations. For example, indie games production organisations were

seen as being more creative than larger multi-nationals. This supports previous research,

such as Perron’s (2003a) which, in reference to women’s participation in the new media

industry in the UK, noted “that work in the larger firms was more likely to become

formulaic and less challenging” (p. 80). Deuze et al. (2007) noted the “tensions” that the

environment presents workers with when their creative freedom is tempered by financial

control (p. 292).

Agency theories can help explain why the agent-driven mechanism of

transforming can influence women’s participation in the DCI. Although not directly

identified by the concepts used from the MTS, creativity is an important aspect of

agency. Agency theorists, such as Bandura (1986), recognised human creativity and

suggested that modelling (observation of a model) could promote creativeness by

exemplifying how to create novel syntheses and fresh perspectives that could challenge

conventional mind sets. Although most research neglects the human agency involved in

creativity, agency itself has a creative element, in the form of creative acts (Sewell, 1992;

Joas, 1997), and imaginative and critical intervention (Emirbayer & Mische, 1998).

Extant literature recognises that creativity can influence participation. Creativity

is seen as a distinguishing feature of the creative class (Florida, 2002), and as being

associated with DCI workers and highly valued in the industry (Gill & Pratt, 2008),

particularly within games production (Deuze et al., 2009). Although most research

regarding creativity in the DCI has focused on the industry rather than the individual,

there is recognition that, for workers in the Australian DCI, receiving feedback on

creative work from their immediate peers and industry peers is enormously beneficial

(Haukka & Brow, 2010). For women in particular, Fullerton et al. (2008) suggested that

an ideal female-friendly job environment encourages creativity, new ideas, interaction

and communication. Similarly, one of the reasons given by teenage girls as to why

multimedia may be more attractive than IT as a career is that the former is seen as

offering the opportunity for greater creativity (Multimedia Victoria, 2007).

Thus, the agent-driven mechanism of transforming is an important influence on

women’s participation; however, further conceptualisation of creativity is required. As

Section 6.3.1 further discusses, the agent-driven mechanisms in the creative act of

agency category can be fostered by strategies that support creative work practices, and

also those that cultivate occupational roles that offer problem-solving opportunities for

workers.

Emotion Ethical Being

5) HUMAN ACT OF AGENCY

The Human Act recognises the human

capacities that an individual possesses. The human act

involves the mechanisms of emotion, ethics, and being.

These mechanisms may be difficult to observe or

predict.

EMOTION

“being passionate...it's a hard one to quantify. I suppose I could get philosophical...” (S1).

As the findings indicate, an individual’s emotions can influence their

participation, and several of the events of interaction highlighted these emotions. From a

Critical Realist perspective, emotions can be rational, intelligent responses to events

(Sayer, 2008).

Agency theories can help explain why the agent-driven mechanism of emotion

could influence women’s participation. Bandura (1997) recognised the salience of

emotions in human agency in the form of an individual’s disposition. Although Giddens’

(1984) modalities did not explicitly identify emotions, his later work (1992) considered

emotions such as fear and trust (p.375), and even passion. Giddens (1992) even

wondered: “Who can live without passion?” (p. 201). However, Meštrovi (1998) argued

that Giddens only “touches on emotions fleetingly, here and there” (p.103). Archer

(2000), whose work aligns to both agency theory and Critical Realism, identified the

salience of a range of emotions, including a sense of trust. Human Agency theories may

offer further insight into emotions.

Further extant literature identifies that emotions can foster a woman’s

participation. Although emotions are not often considered by researchers–being seen as

“dirty work” (Lewis & Simpson, 2007, p. 184), and associated with the private sphere

(Lewis & Simpson, 2007, p. 4)–the literature concerned with women’s participation in

ICT has identified that emotions play a role. Creating ‘fun’ experiences, for example, can

encourage young women into ICT education pathways (Stockdale & Stoney, 2008), with

multimedia being identified by school age girls as being more ‘fun’ than IT (Turner,

2003). In the broader literature, there has been recognition of the need to harness such

emotions to benefit the workplace (Mason 2006, p.9). The Affect Theory literature

(Gregg & Seigworth, 2010) recognises the role of women’s emotions in the workplace.

A central emotion recognised in participants’ accounts is passion, particularly the

passion they held towards their DCI career. It appeared that an individual’s passion was

expressed by adopting practices such as staying back late and spending large amounts of

time playing games, or completing challenging coding tasks at home. However,

participants found it difficult to define passion. This is pertinent, as it was suggested by

two participants (who had hiring responsibilities) that women may not have the same

amount of passion (relevant to the industry) that men may. Passion could fuel

participation in the face of constraining influences, such as poor salary and long hours.

Secondary sources support the idea that emotions, such as passion, play a role in

fostering participation; for example, an email from a UK-based group for women in IT

asks members if they are “feeling passionate and fulfilled with their career/life”

(Appendix 17, Item 26). The email notes that if women do not have their ‘core’ needs

met, they will not be able to focus on their passions.

The emotion of passion has been identified in previous research in the DCI

context (Gill, 2007; Roan & Whitehouse, 2007; Deuze et al., 2009) and the IT context

(Griffiths et al., 2005). Roan & Whitehouse's (2007) study identified passion as being an

important employment trait: “The male senior manager maintained that he hired people

with a ‘passion for games’ and ‘a passion for programming’” (p.29). Indeed, Consalvo

(2008) proposed “one word” as being used to explain why women worked in the games

industry–“passion” (p. 185). As the findings highlight, certain industry practices can

exploit an individual’s passion; this finding supports Consalvo’s (2008) suggestions that

passion leads people into the industry and keeps them going when the conditions are

difficult, and that companies trade off such passion to increase output. Passion is

problematic in that “the ideal worker is constructed as someone possessing a passion for

games, and that passion is used to help maintain work practices that may ultimately kill

the passion” (Consalvo, 2008, p.186).

Findings suggest that, although women see themselves as passionate, they may

not be perceived as being passionate by others. In the extant literature, Vigdor (2011)

recognises that women face challenges in being perceived as being technologically

capability and passionate, describing this as a ‘techno-passion’ gap. Vigdor (2011) also

brings to our attention the role that previous seminal research, from both Camp and

Turkle, has played in perpetuating the notion that women are not passionate about

technology. Lang (2007b) suggested that there is qualitative data that supports a disparity

between men and women in the ICT industry, with respect to passionate interest. Further

research is required into the perceptions surrounding women’s passion for the industry.

The inclusion of emotions as a mechanism does not imply that emotions are the

domain of women. Emotions, such as frustration, surfaced not only for females in the

DCI context, but also for the male freelancer: “...just infuriated me because I’d put so

much work into the soundtrack and then to have it sort of butchered was really

annoying...” (S1). Similarly, in secondary data sources, a male, white Australian games

designer, who responded to the IGDA diversity survey, highlighted that the “school boy

‘conforming’ egotistical and prankster culture” in the Australian games industry led to

the development of his depression and social anxiety. He felt ostracised because he did

not fit in and “it seemed to make no difference to them how much I showed through my

actions that I could design with the best of them and manage a project to success”. He

noted that such a culture was “demanding and taxing emotionally” (Appendix 14, Item

28-M (bi), 25, White, disabled, Uni, Australia). However, ICT-related literature notes

that emotions such as anxiety may influence women and minorities “more often and

more deeply” (Tapia & Kvasny, 2004, p.88). This may be a result of women facing

higher performance pressure because of their increased visibility in the workplace in

which they are a minority (Tapia & Kvasny, 2004). Thus, the agent-driven mechanism of

emotion is an important influence on women’s participation.

ETHICAL

“I don't call that deviant I call that critical thinking” (S2).

The findings indicate that the mechanism of being ethical can influence women’s

participation. Although there were few empirical examples, ethics surfaced as a

mechanism when it influenced certain decisions by participants. One example involved

interpreting a participant’s account of workplace bullying. She described how she

questioned the bullying practices of a manager, or the practices in a workplace that

encouraged secrecy and fear. That she suggested that bullying ‘wasn’t right’ was

interpreted as her questioning the practice.

There was also a specific suggestion that the games industry is perceived as not

being ethical. For example, a stakeholder described how games had been “demonised”

by the media. The games industry is seen as not offering an attractive career path

because of its association with games products that have been highlighted by the media

as perpetuating anti-social morals and content that potentially is not appealing to women,

with violent content being one of the socially problematic aspects of game play (Shinkle,

2008). For example, the computer game Grand Theft Auto has been widely reported in

the media as perpetuating both sexism and violence. One participant questioned playing

violent computer games (which she would potentially be involved in producing) in front

of her toddler.

Agency theories can help explain why the agent-driven mechanism of ethics can

influence women’s participation. Although not included initially as a concept in the MTS,

Bandura’s SCT (1986) suggests morality, and in particular personal morality, is involved

in agency. Similarly, Giddens’ modality of 'Norms' explicitly suggests that an

individual’s action is closely bound up with the moral evaluations people make,

including their awareness of being sanctioned by others. The mechanism of being ethical

could reveal the moral rules that participants are aware of, and that drive their actions.

Further extant literature does not readily identify that ethics may be an important

mechanism. Elliot (2003) recognised that there is a similar sociological aversion to

ethics, with there being an “objectivistically driven fear of emotion and passion” (p. 16).

However, the agent-driven mechanism of ethical appears to be an influence on women’s

participation and requires further research.

“That’s just me...I’m just aware of these things” (M2).

The agent-driven mechanism of being is the most difficult to define, as it has a

metaphysical nature; it includes the humanness of the individual, and traits such as an

individual’s reflectivity. A succinct definition of the agent-driven mechanism of being

presents challenges, and further conceptualisation and empirical data is required. In any

further theorisation of ‘being’, there may be benefit in drawing on thinkers such as

Heidegger. His key work Being-in-the-world (Heidegger, 1962) moves the study’s focus

from a situated everyday interaction to a more ontic or existential focus of his concept of

‘Being-in’.

The most obvious manner by which the mechanism of being surfaced was with

participants’ accounts, which showed a high level of reflectivity with regard to their own

and other people’s actions and circumstances. Indeed, in asking participants about their

experiences in the interviews, there was a sense that they were reflecting or realising

something for the first time. For example, when asked about their future plans for

children, several responded with a long silence; visually, they appeared to be reflecting.

Participants not only reflected on events and influences, but were also reflexive.

Reflection involves the individual considering their own role in an event, whereas

reflexivity involves a higher order understanding, enabling the making of connections

between the current experience and previous situations (Taylor & White, 2000).

Considering this difference between reflection and reflexivity, Archer (2007) suggested

that we need to be precise in “what a subject needs to be aware of in order to be

influenced” (p. 17). Findings reveal that, when participants recounted their experiences

(suggesting reflection), they often described the inaccurate social messages about women

and technology (suggesting reflexivity). Thus, these women may be considered what von

Hellens, Nielson and Trauth (2001) in the ICT literature refer to as “metalevel observers”

(p. 117).

Agency theorists recognise the importance of reflection as a capacity of the

agent. Bandura’s SCT (1986) highlights the causal effect of an individual’s capacity for

self-reflection, and Giddens’ accentuates the “knowlegeability of actors about their social

circumstances” (Archer, 2007, p. 41). An agent is not a ‘cultural dope’ (Giddens,

Duneier, Appelbaum, & Carr, 2008) driven to act by social processes beyond his/her

control and at the mercy of environmental influences. However, “a great deal of social

research just writes out the area of practical consciousness” (Giddens & Pierson, 1998, p.

Further extant literature identifies that being (including reflectivity and

reflexivity) may be an important mechanism. Archer (2000, 2003, & 2007) posited that

reflexivity is the agent’s most important generative mechanisms because, as a subjective

power or capacity of the individual, it mediates structural powers. Reflexivity helps us

“make our way through the world” (Archer, 2007, p. 32), particularly in a world where

globalisation has increasingly freed people from traditional restraints of “common

values” to give the individual a “life of one’s own” (Archer, 2007, p. 32). However, as

Thrift and May (2001) suggested, women face difficulties in finding a space for such

reflection, even when it is an essential part of the self, or of what Giddens (1991) referred

to as ‘self-actualisation’.

Thus, the agent-driven mechanism of being is an important influence on

in the human act of agency category can be fostered by strategies that minimise the

emotional stress women experience in the workplace, due to their gender or parenting

responsibilities.

SUMMARY

This section discussed how the use of the three frameworks (first proposed in

Chapter 3) to analyse empirical data, has helped illuminate the influences on women’s

participation in the Australian Digital Content Industry. Synthesis of the insights from

the frameworks and empirical data led to the conceptualisation of agent-driven

mechanisms. The concept of agent-driven mechanisms recognises that influences

manifest in the interaction between the environment and the individual, where both have

a causal role. However, the emphasis lies in recognising the causal capacity of

individuals. The Five Acts of Agency model encapsulates 10 agent-driven mechanisms,

the conceptualisation of which is supported by the empirical data and extant literature.

These agent-driven mechanisms can help explain women’s participation in the DCI.

There are, however, a number of limitations of the findings that must be

recognised. One such limitation is that there were only a few empirical examples of

certain mechanisms, such as ethics and being. These types of mechanisms are difficult to

observe. Future research could strengthen the findings with further empirical data to

support the conceptualisation of the mechanisms within the Five Acts of Agency.

However, mechanisms cannot always be observed or forced to manifest through

experimental research processes, and there remains the “impossibility of

experimentation” (Yeung, 1997, p.53). Thus, further refinement of the agent-driven

mechanisms will require a continuing balance between empirical data and theoretical

explanation.

There are five acts of agency: 1) enabled, 2) connected, 3) collaborative, 4)

creative and 5) human. Although treated as separate analytical categories in the Five Acts

of Agency model, the acts, and consequently the mechanisms within the acts, are

interrelated. Stage 6 of Danermark et al.’s (2002) model requires identifying the manner

in which mechanisms interact with other mechanisms at different levels, under specific

conditions, as a way of explaining concrete events and processes (p. 111). As Sayer

(1992) reminded us: “Abstract theory analyses objects in terms of their constitute

structures, as parts of wider structures and in terms of their causal powers. Concrete

research looks at what happens when these combine” (p. 116). Hence, the discussion has

involved highlighting some of the linkages between the empirical data and mechanisms,

and among the mechanisms themselves. However, to maintain clarity of discussion,

there is minimal reference to how they integrate. Furthermore, the goal for Critical

Realism is identification of ‘the’ sole mechanism and the elimination of competing

explanations (Yeung, 1997), the identification of the ultimate generative mechanisms

(Sayer, 2000), and the level at which mechanisms manifest. As Danermark et al., (2002)

reminded us, mechanisms manifest across different strata such as atomic, biological and

metaphysical (pp. 61-63).

A limitation of the investigation is the minimal consideration of the linkages,

hierarchy and stratified nature of agent-driven mechanisms. Further research could

consider the relational aspects of mechanisms and ask if the mechanisms identified by

the FAA model are irreducible higher level mechanisms, or if there is a hierarchical

relationship among the mechanisms.

6.2.4 OUTCOME: THE ACTS OF AGENCYTHEORY

This section outlines the logic employed by the researcher to develop an

emerging theory entitled the Acts of Agency. The discussion reflects Stage 6 of

Danermark et al.’s (2002) six stage model, which recommends a synthesis of empirical,

theoretical, and ontological insights through the development of empirically-grounded

theory. The goal of this theory is to offer further explanation of the phenomenon of

women’s participation in the DCI. The theory may also be generalisable to other

contexts of participation and by other cohorts.As Trauth (2011) suggested, “At heart,

a theory is an attempt to understand a phenomenon” (p.5). The theory proposed here is

not comprehensive or complete. Theory can be an “interim product” (Layder, 1998, p.

178), a process rather than a final product (Weick, 1995). The aim is not to suggest

theoretical closure, but rather to present a framework to assist developing a clearer

explanation of complex realities. The emerging theory is not predictive; rather, it

sensitises the researcher to the various elements that may influence women’s

participation.

As Figure 30 illustrates, the emerging theory of Acts of Agency builds on the

models previously presented in this dissertation, including the Sphere of Influence (SoI),

Events of Interaction (EoI), and the Five Acts of Agency (FAA). Each model contributes

to an overall explanation of the influences on women’s participation. Utilising the SoI

sensitises the researcher to the Environment (E) and Person (P) characteristics (Elements

A and B in Figure 30) that can influence women’s participation. The SoI can be thought

of as being a theoretical tool that describes what the context and characteristics of the

influences are. This can be thought of as theory that explains “What is?” (Gregor, 2006).

Utilising the FAA model sensitises the researcher to the possible mechanisms at play and

moves findings towards explaining participation. This can be thought of as theory that

explains “How is?” (Gregor, 2006).

TOWARDS A THEORY OF ACT: ELEMENTS AND RELATIONSHIPS

It is a requirement of theory to specify elements and the relationships among

elements. Accordingly, Figure 30 illustrates the elements and relationships involved in

the emerging theory of Acts of Agency. The element of the Environment (A) requires the

researcher to consider ‘what context participation occurs in’. Recognising the element

Person (B) identifies the person to whom the agent-driven mechanisms apply. Identifying

the Events of Interaction (C) helps sensitise the researcher to possible mechanisms at

play (D). Recognising a person’s actions (E) helps draw further attention to the agent-

driven mechanisms. Recognising action also leads to asking, 'What are the consequences

of the actions?’ (F); that is, ‘What effect do the agent-driven mechanisms have on

participation?’ A deep explanation, based on a focus on mechanisms, can be a major

contribution to theory (Harrison & Easton, 2004). Element (G) requires the researcher to

consider the temporal aspects of participation. Following is a short discussion of each of

the elements in the emerging Acts of Agency theory.

Figure 32: The elements and relationships involved in the Acts of Agency theory

A. ENVIRONMENT

As Figure 30 illustrates, the Acts of Agency theory requires a consideration of the

Environment (E) in which interaction manifests. As Section 3.3.2 discussed, both

Giddens and Critical Realists such as Archer (2007) agree that the environment or

‘structures’ impinge upon the agent. The SoI model is a useful analytical tool for

developing a description of the environment or conditions which, in turn, may give

insight into possible influences. However, the environment does not present ultimate

causes; there is no one constraining “structure” (Connell, 1987, p. 95). The researcher

must consider the uniqueness of the context before asking how the individual interacts

with the environment.

B. PERSON

As Figure 30 illustrates, the SoI model also encourages a description of the

Person (P) element, which acknowledges the characteristics of the individual that may

influence their participation. These characteristics include gender; however, the research

process has carefully considered the labelling of the categories to avoid labels that may

lead to essentialist descriptions of women. There is a need to move beyond simply

labelling a person, and towards understanding what Giddens (1993) called the “process

of interaction” (p. 128).

The emerging theory of Acts of Agency does not label the person category with a

gendered label such as ‘women’. The SoI, for example, uses the label ‘person’, as it is

non-gendered, and the FAA utilises the term ‘agent’ when describing agent-driven

mechanisms (Element D in Figure 30). The term ‘agent’, like ‘person’, may go beyond

the binaries of male and female. However, this investigation is not gender blind; rather,

gender is not a differentiating identifier. The Acts of Agency theory is only gendered in

the sense that the identified conditions are most likely to influence the social identity

labelled ‘women’.

A: Environment

B: Person

C: Interaction E: Actions F: Consequence

(participation)

D: Agent-driven mechanisms

G. Temporality

Sphere of Influence (SoI)

Events of Interaction (EoI)

Five Acts of Agency (FAA)

There is scope for the agent-driven mechanisms in the FAA model to be

applicable to other identities. For example, the mechanism of collaborating may well

foster men’s participation, or the mechanisms of belonging may be relevant to older

workers. Future research could investigate if the mechanisms presented in the FAA model

are pertinent to the participation of other social identities, including gender (males),

ethnicity, and age.

There are several approaches to labelling the person category, each offering a

derivation from the core concept of ‘human being’. Trauth et al. (2004), in their research

on women in IT, emphasised the ‘individual’. Archer (1995, 2000), in her argument for

a nuanced view of human agency from a Critical Realist perspective, proposed that the

individual is a stratified being with emergent properties. She thus distinguished between

the person, the agent, and the actor; and in this distinction, the individual human being is

seen as emergent from (but not reducible to) their biological make-up. Archer (2000)

positioned the agent as “parent” of the actor (p. 11); most importantly, and unlike the

actor, the agent can “reflect on their role positions” (p. 284); the actor is seen as an

impoverished individual, having “no ‘inner passions’ of their own, being only influenced

by society” (p. 129).

Further development of the Acts of Agency theory could refine the person

category to acknowledge these stratified aspects of the person.

C. INTERACTION AND D MECHANISMS

As Figure 30 illustrates, the EoI model draws attention to specific interaction (I)

between the environment (A) and person (B). Identifying events of interaction sensitises

the researcher to the mechanisms (D) at play. The FAA model proposes 10 agent-driven

mechanisms that a woman may harness to influence her participation. These mechanisms

have been previously discussed in detail in Section 6.2.3.

E. ACTIONS

Figure 30 illustrates that mechanisms in the FAA model may become more

readily evident through a person’s actions in situated contexts. The role an individual’s

actions play in their participation has been captured, to some extent, by the labelling of

the agent-driven mechanisms, using terms that imply action. For example, agent-driven

mechanisms use adverbs such as ‘problem solving’.

The Acts of Agency theory draws explicit focus to actions, suggesting that it is

essential to recognise that the actions of an individual influence their participation. The

explanation of how things occur involves recognising that the outcome of an action

follows from mechanisms acting in particular contexts. A focus on action is required to

“understand how individual action impacts on social situations” (Dobson, 2001b, p. 206).

Where interaction is an analytical category, action can be an observable practice.

Mechanisms become evident when individuals, through their actions, display their

personal agency. Thus, recognising actions can help reveal the mechanisms at play.

A focus on action is also supported by the Human Agency theory literature.

Indeed, Giddens (1984) suggested that action and agency are the “same” (p. 51). Agency

must involve a person’s “activity”, and “acting, doing things, making things happen,

exerting power, being a subject of events, or controlling things” (Mills, Durepos, &

Wiebe, 2009, p. 12). To act through action is implicit in agency theory, as “action is an

exercise of human agency, a person's power to act” (Cochran, 1997, p. 28), or to “act

otherwise” (Giddens, 1979, p. 56): “To be able to ‘act otherwise’ means being able to

intervene in the world or to refrain from such intervention, with the effect of influencing

a specific process or state of affairs” (Giddens, 1984, p.14). Without the capacity to ‘act’

or ‘act otherwise’, there would be no human agency.

Agency is “an active constituting process, accomplished by, and consisting in,

the doings of active subjects” (Giddens, 1976, p. 121). This is not to say that meaning

behind the actions is discounted. An individual’s account of actions will involve the

participant’s meaning, as meaning is about the individual’s “symbolic identification”

with the purpose of the action (Castells, 1997, p.7), and meaning is actualised in acts. In

this way, agency is an achieved outcome, and a unique response, from the individual.

F. CONSEQUENCE

As Figure 30 illustrates, the emerging theory of Acts of Agency requires that the

researcher recognise the consequences (Element F in Figure 30) of an individual’s

actions (at times exercised as a choice of inaction) within particular conditions.

Congruent with a Critical Realist perspective, the theory requires the researcher to look

beyond those consequences that the agent recognises. These consequences are tendencies

where “different mechanisms can produce the same empirical result” (Sayer, 1984, p.

108). In this investigation, influences (including the environment and person

characteristics, the interaction of the environment and person, the agent-driven

mechanisms, and the person’s action) led to the consequence of participation.

G. TEMPORALITY

As Figure 30 (element G) implies, there is a temporal aspect to women’s

participation. The SoI model incorporates temporal aspects by including a lifespan

perspective in the social dimension, and the history property in the cultural dimension (as

discussed earlier in Section 6.1.2). Where the SoI has sensitised the researcher to the

temporal aspects that an individual may experience, element G recognises the longer-

term temporal aspects.

Both Critical Realism and Human Agency theory recognise that temporality is

involved in the interaction between the person and their environment. For Critical

Realists, both “historical factors and current context” (Archer, 1998, p.196) are integral

when considering the interaction or interplay between the person and structures.

Similarly, agency theorists also recognise the role of temporality. For example,

temporality was recognised by Giddens with concepts such as the long duree of

institutions and individual lifespan (Calhoun et al., 2002). Temporality can illuminate

why individuals respond in a certain manner, so as to “account for variability between

agents” (Emirbayer & Mische, 1998, p. 963), and to foster recognition of patterns of

tendencies.

Although recognition of temporal aspects can add “richness” to theory

development (Layder, 1998, p. 177), there are challenges to the inclusion of temporality.

One challenge is simply in definition, with one literature review identifying 211 different

conceptualizations of time perspective (McGrath & Kelly, 1986). Future research could

consider temporality by drawing on theorists such as Emirbayer and Mische (1998),

who–in critiquing Giddens for his overemphasis on routine, habitual and taken-for-

granted agency–present three temporal elements of past practices, future possibilities,

and present contingencies. Likewise, Hitlin and Elders (2007) present four types of

temporality within agency–including existential, pragmatic, identity, and life course–and

suggest that a clearer articulation of temporality could aid in understanding the use of the

concept of agency. From the Critical Realist literature, temporality is considered in

Archer’s (1995, 2010) morphogenic model, and in Bhaskar’s (1979) Transformational

Model of Social Action (TMSA), where both acknowledge that structures predate actions.

Future research may refine the element of temporality in the Acts of Agency theory.

GENERALISABILITY OF THE ACTS OF AGENCY THEORY

As noted in Chapter 4, an outcome of good theory is that it is extendable to other

situations (Maxwell, 1992). It is proposed that the Acts of Agency theory and nested

conceptual frameworks (Sphere of Influence, Events of Interaction, Five Acts of Agency)

can be used in other research contexts. Although the Five Acts of Agency model is

context specific (the Australian DCI) and cohort or sample specific (women employed as

interactive content creators) the model itself is agnostic in regards to whom and where

the question of participation is being asked. It is not a model about gender, it simply

recognises that gender is a key sub-category of the person when investigating female

DCI workers. For example, in the Sphere of Influence model, the category of person

could readily be replaced to be another minority cohort such as ethnic (e.g. Asian) and

the context of the environment category could readily be replaced to be an education

context (e.g. a school). It is expected the sub-categories would differ, for example

occupation (as sub-category of social identity of the person) would not be expected be

relevant to a school student. The Sphere of Influence model could provide a starting point

for an investigation into the participation of other minority groups and within other

contexts.

Similarly, the Five Acts of Agency can be used as a starting point to investigate

the mechanisms at play for other identities and in other contexts. It is expected that many

agent-driven mechanisms may be similar, such as ‘accessing’ - the need to access

information, people and technology. The actual nature of the information and technology

may however differ for the specific cohort. It may be expected that certain agent-driven

mechanisms may not be as prevalent in other contexts. For example, the need to be

creative through problem solving and transformation may not be as relevant to nurses in

a hospital context as it is to those working in the Digital Content Industry. However, it is

important to acknowledge that creativity (as a characteristic and action of the individual)

is becoming increasingly important across a range of careers. Further research can aim to

investigate if the mechanisms identified in the Five Acts of Agency model can transfer to

other contexts and cohorts.

To date, early versions of the SoI and FAA models have been used to examine

women’s mobile phone usage in Portugal (See Geneve and Ganito, 2010, and Appendix

1 for all publications stemming from this research). This application of the models is not

necessarily a statement about generalisability, because generalisation is somewhat

incompatible with a Critical Realist stance. It does suggest, however, that the emerging

theory (and analytical models) may be transferable for application to other research

areas.

SUMMARY

This section presented an emerging theory, entitled the Acts of Agency. The

theory builds on the models previously discussed, including the Sphere of Influence, the

Events of Interaction, and the Five Acts of Agency. The discussion has presented the

elements of the theory (including entities such as the environment, person, interaction,

and mechanisms), and relationships (including temporality). There are limitations to the

initial theory, and these limitations may well be addressed through further empirical

work, and also through reference to theorists who can provide insights from a Critical

Realist or agency theory perspective. The emerging theory provides a starting point for

further theorisation regarding women’s participation in the DCI.

Enabled

6.3 APPLICATION OF THE FIVE ACTS OF AGENCY (FAA) MODEL

Chapter 2 proposed that a suitable research outcome would include the

application of findings–that is, an applied outcome. At the beginning of exploratory

research, findings and outcomes are unknown quantities; they emerge as the research

progresses. This section discusses how the findings appear to have an applied value.

Chapter 2 introduced the types of strategies undertaken to address the under-

representation of women in the technology-related industries (See Table 8). Findings

reveal that the Five Acts of Agency (FAA) model provides an empirically derived and

theoretically informed model of concepts against which to evaluate such strategies and,

potentially, to inform new initiatives that can foster women’s participation. This section

discusses how agent-driven mechanisms in the FAA model can both inform

recommendations regarding possible strategies (Section 6.4.1), and offer a way to

evaluate existing initiatives (Section 6.4.2). The mechanisms in the FAA are not

definitive solutions; rather, they illustrate possible approaches for initiatives to foster

women’s participation within ICT and DCI educational and career pathways.

6.3.1 PROPOSINGSTRATEGIES BY DRAWING ON THE FIVE ACTS OF AGENCY

The following are examples of how agent-driven mechanisms can foster

women’s participation. Based on these examples, recommendations for strategies that

would support women’s participation are proposed. A summary of the recommendations

is presented in Table 50.

1) ENABLED ACT OF AGENCY

Women’s participation may be supported by

fostering a person’s capacity for the ‘Enabled’ Act of

Agency. As outlined in Section 6.3.3, the ‘Enabled’ act

involves the key agent-driven mechanism of access; in

particular, access to resources such as people, technology

and information. A women’s participation is fostered when they access resources.

Recommendation 1: Strategies must provide access to resources, such as technology (computers), information and people.

Example: Strategies may include use of the internet to provide access to technology, information and people.

Several participants identified that, in early childhood (the ‘Family’ and

‘Education’ stage in Figure 10); access to resources such as computers and the internet

fostered their skill development. In later life stages (such as ‘Education’ and ‘Work’),

these resources also included internet-based material, which offered access to accurate

industry information, and to tools such as software.

Connected

Imagining Doing

Several participants identified that access to mentors during school and in the

workplace could foster participation. Existing strategies also recommend providing

access to mentors, through the use of internet-based resources as a way to support

women in the DCI. For example, in the international games industry, the IGDA offers an

online mentorship scheme (IGDA, 2005a). Although participants did not name any

particular networking groups, strategies such as Girl Geek Coffee Club increasingly

provide opportunities for the female ‘geeks’ to network both face to face and online100F

2) CONNECTEDACT OF AGENCY

fostering a person’s capacity for the ‘Connected’ Act of

Agency. As outlined in Section 6.3.3, the ‘Connected’

act involves the key agent-driven mechanisms of

imagining and doing. A women’s participation is

fostered when they can imagine their participation, or gain the skills to participate.

Recommendation 2: Strategies must assist women to visualise or imagine their participation.

Example: Promote appropriate female industry role models. A suitable role model is one who emphasises occupational capability, while also acknowledging gender and individuality.

The mechanism of imagining is fostered when participants see suitable role

models. Participants identified the need to see role models they could relate to which, in

turn, could motivate them to work in the industry; seeing role models helped them to

imagine their own participation. However, participants also identified that there were few

role models for women in the industry, in particular, few role models of working

mothers. Thus, strategies should aim to make role models available and visible to both

girls in the education stages, and women working in the workforce.

The appropriateness of these role models must be considered. It is recommended

that role models emphasise their capability as professionals, rather than their gender.

Strategies that emphasise a sexualised gender must be minimised.

One specific strategy that aims to provide role models for the ICT context is the

2006 IT Screen Goddesses calendar95F

98. The calendar presents images of women who work

in the ICT industry. Biographies in the calendar suggest the women have appropriate

skills for the industry. However, the images present the women in provocative poses,

based on popular movies; for example, an Oxford graduate, an Associate Professor, and a

grandmother pose as the actress Sharon Stone in a sexualised scene from the movie Basic

97 Geek Girl Coffee Club began as an initiative in Queensland (Australia) in 2009, and has a continuing growing Facebook presence. 98 Website of IT Screen Goddess calendar http://www.itgoddess.info/

Instinct. Proponents may suggest that this initiative reflects a post-modern discursive

playfulness in the re-appropriation of imagery, as a tool to challenge the discourse that

there are no women working in ICT. Nevertheless, concerns were raised about its

appropriateness, as the following quote suggests: “A calendar of scantily clad women

definitely does not attract this female to IT. Show me challenging work, a career path,

good working conditions, respect and The Money, and I would consider it” 96F

Consequently, the sponsorship from the Australian Computer Society (ACS) was

withdrawn upon publication. The ACS president stated: “It is untenable for us to be

portrayed as supporting a publication with a naked woman on the cover, in the name of

improving the image of women in IT”100 Findings also reveal that the use of booth babes

at games-related conferences was seen as a deterrent for women: “More young girls

would aspire to work in video games if they could imagine themselves as more than

exhibition props. Video game companies should support that imagination in anyway

feasible” (Huntemann, 2007).

Thus, sexualising women working in the DCI or ICT industry is seen as

promoting inappropriate role models.In contrast, strategies aiming to provide role

models to young girls (primarily aged between 8-14), such as Tech Girls are Chic: Not

Just Geek (Appendix 17, Item 50), provide industry information alongside insights into

more stereotypically feminine pursuits, such as fashion interests. However, in contrast to

the IT Screen Goddess calendar, content is not sexualized; this is particularly salient

considering the younger target audience.

In addition, strategies promoting role models may consider the use of the word

‘geek’. The term ‘geek’ has been widely reported as having negative connotations,

particularly for young girls (Beekhuyzen & Clayton, 2004). Strategies such as the Sexiest

Geek Alive 98

101 pageant, which recognises that IT is “not one of the sexier

professions”,102aim to redefine the ‘sexiness of geekiness’ for both male and female

participants by focusing on “brains, personality and technical creds”, rather than “looks”.

The term ‘geek’ can also suggest expertise in technology. Strategies promoting ‘geeky’

women, such as Girl Geek Dinners, Girl Geek Coffee Club (See Appendix 3), and ‘She’s

Geeky’ (See Appendix 17, Item 60) intimate that women can be experts with regards to

technology. Such strategies contrast existing gender stereotypes, which perpetuate

women’s limitations regarding their technical capability.

Recommendation 3: Strategies must foster women’s connection to the industry image.

99 Helen July 12, 2006 05:24 PM The Sydney Morning Herald Blogs: MashUp / Cheesecake calendars Archives 100 Comments reported on third party blog http://rosemary.id.au/view/blog/hullabaloo/ 101Sexiest Geek Alive pageant http://archive.salon.com/tech/log/2000/03/07/sexiest_geek/index.html 102 As reported online in popular media http://news.ninemsn.com.au/article.aspx?id=284665

Example: Articulate an accurate industry image to highlight the various occupational roles involved in the DCI; challenge outdated stereotypes that the industry is all about programming.

The mechanism of imagining is fostered when participants can visualise their

involvement and develop accurate outcome expectations. Recommendation 1 notes that

women need access to industry information. Recommendation 3 emphasises that this

information must be accurate, as it helps inform women’s perceptions of the industry

and, ultimately, their goals.

Findings suggest that the unappealing image of the DCI and related ICT industry

can influence women’s participation. Secondary data also suggests that stereotypes

surrounding ICT result in the perception that workers in the industry are “boring,

difficult, and antisocial” (Appendix 17, Item 59 ). IT lacks appeal as a career as it is seen

as boring (Timms, Courtney & Anderson, 2006; State of Victoria, 2007, p.14). The

image of certain roles, in particular programming, is perceived as unattractive by young

people (Rettenmayer, Berry, & Ellis, 2007). Although the DCI does inherit some of the

connotations associated with the ICT industry, it is seen as more appealing, as roles are

also associated with creativity (Gill, 2002; 2007). However, this more appealing image is

not always communicated by the DCI industry.

Strategies must present accurate and detailed descriptions of occupational roles

and skills. If there is any ambiguity in job descriptions, women may believe that they do

not have the required skills, or are placed in a position where their outcome expectations

are not met. Inaccurate information may in part contribute to the expectation gaps that

Tapia and Kvasny (2004) noted for women in the ICT industry. Furthermore, there is a

general belief that the DCI requires workers to work long hours. However, findings also

reveal that not all roles face such demands. For example, one participant suggested that

most artists went “home by 5.30pm” (G1). Similarly, several participants suggested that

programmers worked longer hours than those in other roles. However, this was not the

actual experience of the programmers themselves. Thus, the industry must actively

address inaccurate information surrounding occupational roles and work practices.

Recommendation 4: Strategies must offer women support to develop skills relevant to the industry, including ‘technical’ skills. Example: Strategies must offer training to women, both early on at school, and later in the workforce.

The mechanism of doing involves women developing skills relevant for DCI.

When participants developed industry-relevant skills, this helped them improve their

self-efficacy and to challenge gender stereotypes of women’s technical proficiency.

These skills could be developed in a variety of ways; for example, by making personal

websites: “start off with a little personal website...” (M9). The participants developed

Collaborative

Belonging Sharing

skills through formal training (such as the Diploma of Multimedia for Women),

workplace training, or self-directed training. Work experience was also seen as providing

an opportunity to develop industry relevant skills: “Work experience is really good too; if

you didn’t want to commit to a full course initially then you can get a little bit of a feel for

how a role is” (M9). However, work experience opportunities appear to be limited in the

DCI. A similar situation has been noted by Walters (2006) for IT students. Thus,

initiatives aiming to encourage women’s participation must offer training opportunities

across education, workplace, and personal contexts.

Recommendation 5: Workplace strategies can support a continued connection by the female DCI worker.

Example: Provide opportunities for skill maintenance, particularly for working mothers, through flexibility in work practices.

The agent-driven mechanism of doing is fostered when participants not only

develop, but also, maintain their skills. Participants identified that they faced challenges

in regards to staying in the industry once they were employed. Participants needed to

develop skills and maintain those skills in the light of ever-changing technology (for

example, new technology), and personal circumstances (for example, when becoming a

parent). Those participants that had children described actions that they took to maintain

their skills while also addressing their parental responsibilities, such as ensuring they had

software at home when on maternity leave.

The ICT literature suggests that the gender gap in the IT workplace can be

reduced by introducing flexible schedules, providing refresher training for workers who

temporarily leave, and accommodating work/family balance (Lanzalotto, 2007). Flexible

working initiatives would allow more women to remain in the sector; however, such

initiatives must not result in lower pay and status, or diminished promotional

opportunities (DTI, 2005). Australian DCI organisations must provide avenues for

women to maintain their skill currency.

3) COLLABORATIVE ACT OF AGENCY

fostering a person’s capacity for the ‘Collaborative’

Act of Agency. As outlined in Section 6.3.3, the

‘Collaborative’ act involves the key agent-driven

mechanisms of belonging and sharing. A woman’s

participation is fostered when they collaborate and share experiences with other people.

Recommendation 6: Strategies must foster a women’s sense of belonging within the industry. Example: Provide a work environment that fosters inclusion of minority workers.

The agent-driven mechanism of belonging may manifest in environments where

people feel supported. Participants described the importance of fitting into the team

environments of the DCI workplace. Agosto et al. (2008) suggested that, in an ICT

educational context, social cohesion and peer support fosters women’s participation.

However, both the findings from this investigation and the extant literature identify that

being in the minority, due to gender ratios in the workplace, can leave women feeling as

though they don’t belong. If women are in the minority, organisations should be sensitive

to the need to foster a sense of belonging. For example, organisations could provide

collaborative workspaces and communal lunch areas.

Recommendation 7: Strategies must foster a culture of sharing in the workplace. Example: DCI organisations should incorporate collaborative tools within the work environment, and use these technologies to advertise employment opportunities.

The mechanism of sharing is fostered when participants can readily share ideas,

knowledge, and experiences. Online technologies foster collaboration; indeed, their

influence over the last decade may have already supported a change in women’s

participation: “I think things already have been changed because people are using

Facebook, MySpace, Digg whatever it might be” (M7). Participants described using

online websites to find information that assisted them in their roles; for example, a

programmer described finding useful examples of code. Online tools can facilitate the

sharing of information, resources, and experiences among DCI workers. These tools

could supplement traditional avenues of communication, such as face-to-face

communication, that are integral to the process of accessing job opportunities.

Recommendation 8: Strategies must foster a woman’s sense of belonging to the workplace culture. Example: Organisations must clearly state inclusive organisational cultural practices.

The mechanism of belonging may align to previous research that suggests that a

sense of cultural fit within the workplace is important for women. Previous research into

women in IT has noted the relevance of a sense of fit (Tapia & Kvasny, 2004). Findings

reveal that certain unprofessional male-orientated practices in the DCI can ostracise

women, rather than foster a sense of belonging to the organisational ‘culture’. Examples

of these practices include the sexualised booth babes, and meetings conducted in strip

clubs.

To foster a sense of ‘fit’, DCI organisations should have clear and professional

work practices, including a description of an organisational culture that fosters equity

among workers. This ‘culture’ may need to cater to younger workers who require family-

friendly policies, as well as to an ageing workforce. Even something that seems a small

aspect of workplace culture can have an influence. For example, by having a relaxed

Creative

Problem-solving Transforming

Emotion Ethics Being

dress code, individuals have the flexibility to express their individuality, and women are

not forced to conform to gendered norms of wearing dresses and skirts.

4) CREATIVE ACT OF AGENCY

Women’s participation can be supported by

fostering a person’s capacity for the ‘Creative’ Act

of Agency. As outlined in Section 6.3.3, the

‘Creative’ act involves the key agent-driven

mechanisms of problem solving and transforming. As noted on page 274, creativity is an

integral aspect of the industry, and appears to be a key motivator for its workers. A

woman’s participation is fostered when she can be creative.

Recommendation 9: Strategies must foster opportunities for women’s creativity. Example: Organisations should provide individuals with opportunities for challenging work roles.

The mechanism of problem solving can be fostered in the DCI context by

providing challenging work to workers. Participants explained that being able to problem

solve was a motivating aspect; for example, the challenge of a coding problem or

addressing client requirements. Related research recommends fostering problem solving

approaches in ICT education, suggesting that it is preferred by female IT students

(Agosto et al., 2008). Interventions have recognised that encouraging young girls’

creativity could foster their participation (Lang, Craig, Fisher & Forgas, 2010). Thus,

DCI organisations must provide workers opportunities for challenging work roles,

possibly through research and creative play.

5) HUMAN ACT OF AGENCY

Women’s participation can be supported by

fostering a person’s capacity for the ‘Human’ Act

of Agency. As outlined in Section 6.3.3, the

‘Human’ act involves the key agent-driven

mechanisms of emotions, ethics, and being. A

woman’s participation is fostered when she feels a sense of completeness.

Recommendation 10: Strategies must recognise the role of emotions. Example: Organisations can offer environments that foster positive emotions.

The mechanism of emotion surfaced in participants’ accounts; for example,

participants recounted experiencing positive emotions, such as fun, while playing games.

Existing strategies aiming to foster young girls’ participation, such as Technology Takes

you Anywhere, may foster their positive feelings about technology as the initiative aims

to provide a fun experience.

Participants also described experiencing constraining emotions in the workplace

context. For example, a culture of working long hours may leave working parents feeling

guilty for leaving work earlier (due to family responsibilities) than their colleagues.

Although both genders have parenting responsibilities, women may face greater gender-

related stereotypes as a result of their parenting responsibilities. There were also positive

emotions in the workplace, such as pride: “I noticed that Queensland seems to take pride

in the games industry and it also tries to encourage women” (G7). Thus, strategies that

foster positive emotions through the promotion of personal and industry-level success

can foster participation.

Recommendation 11: Strategies must recognise the role of ethics. Example: The DCI can more actively promote the ethical products that it produces.

The mechanism of ethics can be fostered if the DCI illuminates the different

types of products and practices that exist, such as those emerging from serious games

development. Studies indicate that young women see other disciplines, such as education

or psychology, as leading to careers that are more focused on solving human problems;

consequently, they find these fields more attractive in comparison to IT (Agosto et al.,

2008). There is some indication that ethics is surfacing as a consideration for the

industry, particularly in game design (Takahashi, 2010). However, there are few

strategies to directly promote ethics as an aspect of the industry, and little research (in the

DCI context) regarding ethical decisions made in product development (Thompson &

Smolt, 2001), or the ethics of games content (Sicart, 2009). One initiative that has drawn

attention to the salience of ethics for IT workers is the ITMillion web site (See Appendix

17, Item 58), which asks people to share their reasons for pursuing an ICT-related career.

Respondents to the website emphasise that doctors and lawyers ‘help people’, and this

suggests these careers are perceived as having societal value.

Strategies could include promoting the various ‘ethical’ products developed by

the DCI; for example, serious games such as Darfur is Dying have attracted praise for

addressing social issues.101F

103 As Fullerton et al., (2008) suggested, a focus on the social

value of games could attract women to the industry. This perspective is consistent with

Margolis and Fisher’s (2002) research, which indicated that female computer science

majors are motivated to help the world, rather than simply learn algorithms.

103 http://www.darfurisdying.com/ (Last accessed 30 November 2008)

Table 50 A summary of recommendations to foster the participation of women in the DCI Recommendation 1: Strategies must provide access to resources, such as technology (computers), information, people. Recommendation 2: Strategies must assist women to visualise or imagine their participation.

Recommendation 3: Strategies must foster women’s connection to the industry image.

Recommendation 4: Strategies must offer women support to develop skills relevant to the industry, including ‘technical’ skills.

Recommendation 6: Strategies must foster a women’s sense of belonging within the industry.

Recommendation 7: Strategies must foster a culture of sharing in the workplace.

Recommendation 8: Strategies must foster a woman’s sense of belonging to the workplace culture.

Recommendation 9: Strategies must foster opportunities for women’s creativity.

Recommendation 10: Strategies must recognise the role of emotions.

Recommendation 11: Strategies must recognise the role of ethics.

RELATIONAL ACTS AND STRATEGIES

Although Table 52 appears to suggest that the mechanisms in the FAA model–

and, consequently, the possible strategies–are not relational, there is considerable

overlap. For example, being able to access technology (Enabled act) may, in turn, help

women feel connected to the industry (Connected act) through developing skills valued

by the industry. A confidence in their skills may foster women’s capacity to feel like they

belong in the work environment (Collaborative act). Skills may also underpin their

capacity to be creative (Creative act) and, ultimately, the mechanisms of emotions

(Human act) such as pride. Thus, the FAA model recognises that several mechanisms

may be simultaneously at play, and encourages an understanding of influences, and

ultimately of strategies, in a relational manner.

Table 51 further illustrates the relational aspect by highlighting how all Five Acts

of Agency are evident when explaining the influence of ‘Motherhood’–one of the 10

Events of Interaction identified in the findings. The findings reveal that women need

access to information to make decisions regarding maternity leave options. Women faced

difficulties in staying connected with the industry as there were no role models working

in the industry; thus, they could not imagine their continued participation as working

mothers. Women’s capacity to be collaborative (that is, to feel as though they were ‘one

of the boys’) could also be challenged if they were the only team member to take

maternity leave. While on maternity leave, women could remain connected by

maintaining skills. The difficulties women face in balancing work and life commitments

could lead to their facing emotions of guilt. Such emotion prompted them to reflect on

the value of staying in the industry. An event such as becoming a parent also illuminates

the temporal influences. For example, it has only been in very recent history that

Australia implemented a national paid parental leave scheme8; this may provide women

with resources and options they previously have not had. 104

Table 51 The Five Acts of Agency and strategies for the event of motherhood The Five Act of Agency as evidenced in the empirical data

Possible strategies

Enabled -Limited access to adequate information regarding maternity leave constrained participants’ informed decisions

Government and industry to provide career information on maternity leave options as a part of career planning; HR to provide information, for example, suitable childcare options

Connected- A lack of working mother role models

Provide images of working mothers; however, also recognise that working mothers need to retain their professional identity

Stepping away from the work environment may be detrimental to maintaining skills

Provide access to resources that foster skill development when on maternity leave; for example, access to conferences, software and hardware

Collaborative -Working mothers can feel they are the ‘odd ones out’ in the workplace setting

Foster communication among working mothers across organisations, so that they can share information and advice

Working mothers can feel they ‘out the loop’

Provide ongoing contact with those on parental leave (e.g., through email); give the opportunity to ‘drop by’

Creative -Working mothers face changes that may influence their ability to remain creative

Creativity or the need to be creative does not disappear with motherhood; recognise that it continues and may transform into other skills, ideas

Human -Motherhood will place demands on a person’s emotions

Allow some flexibility in work practices so that women can maintain their passion; do not foster environments that create a sense of guilt (for example, when leaving work early for parental purposes)

6.3.2 EVALUATING STRATEGIES BY DRAWING ON THE FIVE ACTS OF AGENCY

The previous section presented several proposed strategies to foster the mechanisms in the

FAA model. Table 52 summarises several empirical examples of the mechanisms in the

FAA model that are linked to the recommended strategies to foster women’s participation.

104 An example of contemporary influences includes the fact that Australia was the second last OECD country to implement a national plan for parental leave (in 2011).

Table 52 Examples of agent-driven mechanisms (in the FAA model), linked to possible strategies Act/ Mechanism

Evidence (for example, individual action) of how mechanisms influenced participation

Linkages to environmental characteristics

Recommendations to foster participation

1. Enabled (Access)

Seeking industry information on web, and contacting industry mentor, helped individual make informed decisions

Limited accurate industry information

Provide accurate industry information, and networking events, e.g. via an industry website

2. Connected (Imagining)

Observing role models helped the individual set goals and future plans; for example, having children

Lack of appropriate female role models

Promote women in industry in non-essentialist manner; for example, not as ‘sexualised booth babes’

Connected (Doing)

Skills practice and training fostered individual confidence and team ‘fit’

Stereotyping lowers women’s self-efficacy

Challenge stereotypes; provide access to training and skill development

3. Collaborative (Belonging)

Offering assistance to others, fostered individual’s sense of ‘fit’ and feeling supported: “We’re like a group of friends”

Male majority leaves women feeling ‘different’

Foster socially inclusive workplace practices; for example, shared eating areas

Collaborative (Sharing)

Sharing skills fostered intrinsic reward: “People ask me…”

Social stereotypes regarding women; ‘soft skills’ biases

Provide mentoring schemes (through social media technology)

4. Creative (Problem solving)

Seeking challenges fostered intrinsic reward through meeting these challenges

Rapidly changing technology leads to burn out

Implement workplace management strategies; for example, time for training

Creative (Transforming)

Fulfilled creative potential; fostered intrinsic reward and recognition

Pressure to develop ‘saleable’ product constrains experimentation

Implement workplace management strategies to ensure ‘play’, research

5. Human (Emotion)

Passion fuelled effort and motivated the individual

Social stereotypes regarding women’s passion

Encourage work practices to manage emotions; for example, stress, guilt over childcare, passion

Human (Ethical)

Observing social stigma surrounding certain games prompted the individual to question their involvement

Games content diversified recently (for example, Wii)

Highlight ethical products and practices within the industry

Human (Being)

Fostered future planning: “I haven’t thought about that”

*Link to recommendations identified in Section 6.3.1 **no strategy proposed

In addition, the FAA model can help evaluate existing strategies by asking if and

how the strategies foster the agent-driven mechanisms. For example, the

recommendations put forward by the Smart Women – Smart State Strategy Science,

Engineering and Technology Action Plan 2006 – 2009 (Queensland Government, 2006c)

(See Table 53) can be understood with reference to the mechanisms in the FAA.

Recommendation 1 in the action plan recommends ‘promoting the use of accurate

terminology’. Promoting accurate information can be seen as fostering a person’s

capacity to access accurate information, thus aligning to the agent-driven mechanism of

‘accessing’. As Table 53 shows, all the recommendations presented by the action plan

can be evaluated by asking if and how they foster the agent-driven mechanisms in the

FAA model.

6.3.3 SUMMARY

This section discussed the application of the Five Acts of Agency (FAA) model.

The 10 agent-driven mechanisms within the model can be of assistance when planning or

evaluating strategies and initiatives that aim to promote women’s participation in

technology-related careers.

Section 6.4.1 recommended 11 strategies to foster women’s participation on the

basis that they foster agent-driven mechanisms in the FAA model. For example, women’s

participation may be fostered if a strategy supports a women’s capacity for the agent-

driven mechanism of accessing. The empirical evidence supports the conclusion that

access to resources such as technology, information, and people can influence a women’s

participation in the DCI. Women may not participate if they face obstacles in accessing

resources; for example, if a ‘boys club’ culture in the DCI workplace prohibits women’s

access to information.

Section 6.4.2 proposed that existing strategies can be evaluated against the FAA

model. Existing strategies–such as those from the Smart State Strategy (Queensland

Government, 2006c) can be understood by drawing on the concept of agent-driven

mechanisms. For example, the strategy recommends role models which, findings

suggest, supports a women’s capacity for the agent-driven mechanisms of imagining.

The FAA model provides an original approach to recommending and evaluating

strategies that may foster women’s participation in the DCI. Further work remains in two

key areas. First, in the collection of further empirical data to ascertain if, and how, the

agent-driven mechanisms are relevant to strategies. Second, there is a need for further

consideration of the agent-driven mechanisms in a holistic manner, as findings to date

suggest that a combination of mechanisms may be at play. For example, when a woman

becomes a parent, several agent mechanisms may influence her participation. Thus,

strategies need to consider not only how they foster an agent-driven mechanism, but also,

the value of fostering several mechanisms.

Table 53 Use of the Five Acts of Agency model to evaluate Smart Women – Smart State Strategies Science, Engineering and Technology Action Plan, 2006 – 2009 (Queensland Government, 2006c)

Smart State Smart Women Strategy Act of Agency (agent-driven mechanisms)

1. Promote the use of the terminology ‘science, engineering and technology’ in Queensland schools • Investigate the branding of resource materials developed by the Queensland Studies Authority and the Department of Education and the Arts to promote the science, engineering and technology areas, utilising language and ideas attractive to girls

Outcome: Increased exposure to Science, Engineering and Technology as distinct disciplines for girls in the early and middle years of schooling Measure: Progressive adoption of ‘science, engineering and technology’ terminology in resources produced by Education Queensland

Enabled (Access- resources such as information) Connected (Imagining)

Providing accurate information assists in making informed career decisions Industry presented in an appealing manner; clarifying types of roles minimises ambiguity and a sense of risk, and can assist goal establishment

2. Utilise existing Queensland Government awards programs to increase the recognition and visibility of women in science, engineering and IT careers • Include a category for the recognition of female engineers in the Smart Women-Smart State awards

Outcomes: Recognition of the contribution of women in SET fields Increased role models for women in engineering Measure: Number of awards programs to promote women in SET fields

Connected (Imagining)

Role models help form goals and motivations

3. Undertake a feasibility study for the establishment of a SET Resource Centre for women and girls • Could provide a gateway to high quality information and resources on SET careers for women and girls in Queensland

Outcome: Proposal for a “one-stop-shop” on SET resources and careers for women and girls in Queensland submitted to Cabinet Measure: Proposal developed by the Smart Women – Smart State Taskforce

Enabled (Access- resources such as information)

Information can help establish goals by demystifying the industry and presenting career options

4. Work with universities and Engineers Australia to develop a Women in Engineering workshop program which brings together female engineering students from across Queensland • To be convened by Queensland universities on a rotational basis

Outcomes: Development of a cohort of female students for research and peer support Decrease in the attrition rate of women from engineering courses who attended the workshop Decreased sense of isolation for female engineering students who attended the workshop Measures: Percentage of workshop participants who report a decreased sense of isolation (three months after the workshop) Attrition rate of workshop participants compared to non-workshop participants

Collaborative (Belonging, Sharing)

A sense of belonging can be fostered by providing access to other women; less likely to be sanctioned in a supportive environment Potential for sharing skills and experiences through collaborative work

5. Identify the range of current SET programs within Queensland that target girls in schools and promote effective models to Government and Industry May include: • Careers Expos • SET in residence programs and • Secondments of industry professionals into schools and vice versa

Outcome: Increase in the number and range of promotional activities for girls in SET Measure: Number and range of programs identified and new ones initiated

Enabled (Access to information), Connected (Imagining, doing)

Enable through information Connect by providing role models Connect by fostering skills development

6. Work with industry to identify best ways to promote family friendly employment practices in SET industries • Promote work/life balance tools • Highlight the Equal Opportunity for Women Agency’s (EOWA) annual awards where award winners are from SET backgrounds • Work with Engineers Australia to expand their industry awards program to promote greater recognition of family friendly employment practices

Outcome: Individuals and organisations involved in SET have access to information on how to implement family friendly employment practices Measure: Number of occasions where work/life balance tools and/or information are promoted

Connected (Imagining)

Foster connection by assisting women to imagine future work goals Minimise the conflict between work and family commitments by providing examples of family friendly practices Motivate participants through rewards such as awards program

7. Promote opportunities for women in SET industries to take up leadership roles • By targeting women in SET careers and promoting upcoming Board opportunities • By encouraging professional development and coaching opportunities for women in SET careers preparing for senior management positions

Outcome: Increased number of women from SET backgrounds on Queensland Boards Measure: Number and range of development and leadership opportunities available to women from SET backgrounds

Enabled (access to resources such as people), Connected (Imagining)

Foster a sense of connection by providing female leaders who are role models or mentors

8. Provide information to assist women re-entering the SET workforce after a career break • Explore and promote networking and mentoring opportunities • Link women to information, both before and after career breaks, to assist them with career planning

Outcome: Skilled women seeking to return to a career in science, engineering or technology have access to information about getting support, developing their skills and making contacts Measure: Number and type of networking opportunities and information provided

Enabled (Access- Information and People)

Enable by providing women access to information and people

9, 10, 11. Final three strategies are project-orientated rather than participation-orientated

6.4 CHAPTER SUMMARY

This chapter discussed the empirical, theoretical, methodological and explanatory

contributions the research provides in regards to understanding the research problem:

‘Why do women participate in the Australian Digital Content Industry (DCI)’.

The research addressed Research Question 1, which asked: ‘What are the

influences on women’s participation?’ Application of the three frameworks proposed in

Chapter 3 revealed a diverse range of influences, which include the characteristics of the

environment and person, events, and underlying mechanisms. Each of the three

individual frameworks helps to identify influences.

The descriptive insights emerging from the application of Framework 1

(analytical framework) illuminates the characteristics of the environment, the person, and

their interactions that may influence women’s participation (See Section 6.1.2). The

refinement of the categories in Framework 1 through analysis of empirical data, led to

the development of two models: the Sphere of Influence (SoI) and Events of Interaction

(EoI) (See Section 6.2.2 and 6.2.3 respectively). Together, these two models foster a

unified description of the environment and person characteristics, and the events that

may influence women’s participation in the DCI.

Employing Framework 2 (theoretical framework) extended the description by

drawing on concepts from the Human Agency theories (proposed in Chapter 3). Section

6.1.6 discussed the value in employing multi theories to scaffold the sensitisation,

analysis and theorisation process. Where a Multi Theory Scaffold, in general, offers

research many benefits; a Human Agency Multi-Theory Scaffold offers a useful specific

theoretical perspective for understanding the empirical data and research phenomenon

under study.

Framework 3, which was based on a Critical Realist ontology (in general) and

Bhaskar’s ‘three domains of reality’ (in particular), fostered the identification of

underlying mechanisms. This recognition of mechanisms, as a part of a stratified reality,

is central to Critical Realism. Thus, all three frameworks reveal different types of

influences.

Furthermore, the research has addressed Research Question 2, which asked:

‘How may we understand these influences?’ In response, the chapter has discussed how

each of the frameworks helped foster an understanding of the influences. Framework 1,

the analytical framework based on existing research, helped to organise empirical data

and link the empirical data to previous research. Framework 2 began to offer further

explanation of influences. Where Framework 1 drew attention to influences such as a

lack of female role models, Framework 2 and, specifically, Bandura’s SCT concept of

self-efficacy (See Bandura, 1986; Lent et al., 1994) explains that same-sex role models

can foster self-efficacy. Human agency theory also emphasised the causal capacity of the

person. Framework 3 moved findings further towards explanatory outcomes by fostering

a causal explanation, with the emphasis on the recognition of underlying mechanisms

and causal tendencies. Thus each of the frameworks provides a way to understand the

influences on women’s participation.

However, as Section 6.3 discussed, it is the synthesis of the findings emerging

from the application of each one of the three frameworks that results in the key

contribution of the research. This synthesis resulted in the conceptualisation of 10 agent-

driven mechanisms within a model entitled the Five Acts of Agency (FAA) (as discussed

in Section 6.2.3). The 10 agent-driven mechanisms are: 1) accessing, 2) imagining, 3)

doing, 4) belonging, 5) sharing, 6) problem-solving, 7) transforming, 8) emotion, 9)

ethical, and 10) being. It is proposed that agent-driven mechanisms provide a way of

understanding why women participate in the Australian DCI. This chapter illuminated

these mechanisms by linking the conceptualisations to further empirical evidence and

extant theory.

Section 6.2.4 extended the synthesis by presenting an emerging theory, entitled

Acts of Agency; this theory integrates the three models arising from the research: 1)

Sphere of Influence (SoI), 2) the Events of Interaction (EoI) and the 3) Five Acts of

Agency (FAA). The theory provides a fruitful position for further theoretical elaboration

in future research.

A further contribution of the research was offered in Section 6.4, which

illuminated how the agent-driven mechanisms within the Five Acts of Agency model can

be utilised when planning or evaluating strategies and initiatives which aim to foster

women’s participation. Eleven recommendations to foster women’s participation were

made in reference to the agent-driven mechanisms, and are supported by empirical data

and existing research.

In summary, this chapter discussed the findings emerging from the analysis of

empirical data using three frameworks: 1) analytical, 2) theoretical, and 3) ontological.

The synthesis of findings results in several original contributions, including models and

an emerging theory. The following chapter provides a conclusion to the research. As

Morgan and Smircich (1980, p.12) avow: ‘Such conclusions depend very much on the

frameworks through which data is interpreted.’

Chapter 7. Conclusions ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 302 of 401

Chapter 7: Conclusions

The research presented in the previous chapters provides a contribution towards

understanding the research problem: ‘Why do women participate in the Australian

Digital Content Industry (DCI)’. The need to address this research problem arises

because women are under-represented in the DCI. The research addressed two key

research questions, which were: 1) What are the influences on women’s participation in

the DCI, and; 2) How might we understand these influences. These questions were

addressed by employing three frameworks to analyse a volume of empirical data

collected as a part of an exploratory case study; the primary data included interviews

from women working in the Australian DCI.

Chapter 2 discussed the previous literature concerned with women’s participation

in both the Australian and international DCI and the related Information

Communications and Technology (ICT) industry. The literature review identified a wide

range of possible influences on women’s participation. As Chapter 3 explained, these

influences could be categorised into two broad categories: 1) Environment, and 2) Person

and the interaction between these two categories required more consideration than was

evident in the existing body of multi-disciplinary literature.

The literature review also highlighted methodological issues, such as the need for

both empirical data and theoretical insights from a strongly pluralistic approach to

understand the complexity of the influences on women’s participation. Accordingly,

Chapter 3 then presented three frameworks (as listed below) with which to analyse the

empirical data. The conceptual commonality between the frameworks is their focus on

interactions between the environment and person. However, each framework also

allowed for the identification of different influences and different interpretations of the

influences on women’s participation.

1. Analytical framework–based on existing research and comprised of three

categories, 1) Environment, 2) Person, and 3) Interaction (proposed in Section

3.1, and discussed in 6.2.1)

2. Theoretical framework –based on a Multi-Theory Scaffold (MTS) and

comprised of four Human Agency theories (HAMTS) which provide 11

concepts for analysis (proposed in Section 3.2, and discussed in 6.2.5) and

3. Ontological framework –based on Critical Realism and provides three

categories; specifically, Bhaskar’s three domains of reality: ‘empirical, actual,

and real’ (proposed in Section 3.3, and discussed in 6.2.6)

Chapter 4 described the research design, including the exploratory case study

approach adopted as the most appropriate strategy of enquiry to investigate the research

questions. The case study methodology offered a way of combining empirical,

theoretical and ontological insights to deliver descriptive and explanatory outcomes.

Chapter 5 presented the findings that arose from sequentially employing each one

of the three frameworks to analyse the empirical data. The application of these

frameworks successfully revealed both influences previously identified in related

literature, and new influences pertinent to the DCI context.

Chapter 6 discussed the key influences that arose from examining the data

through the frameworks, both individually and collectively, and synthesised the

influences into a series of models: 1) Sphere of Influence (SoI), 2) Events of Interaction

(EoI), and 3) the Five Acts of Agency (FAA). Each model reveals a range of different

types of influences. Influences can include the environment and person characteristics

(as encapsulated in the SoI); they can include specific events (as encapsulated in the EoI),

the agent’s cognitive processes (as highlighted by Human Agency theory), and

underpinning mechanisms (as encapsulated by the Five Acts of Agency). The

identification of the influences women face helps foster an explanation of why women

do, or do not participate in the DCI.

Chapter 6 also considered the research value in applying the frameworks and

suite of theoretical models (as will be further explicated in Section 7.1.2). Although each

framework reveals different influences, their real value was shown to be in their

integrated use to foster the development of the conceptualisation of agent-driven

mechanisms. Therefore, the key discovery presented in Chapter 6 was that the most

fruitful research outcome was the combination of the insights from the application of the

three frameworks, which led to the conceptualisation of agent-driven mechanisms (as

was discussed in Section 6.2.3). This conceptualisation was explicated as the Five Acts of

Agency model (comprised of 10 agent-driven mechanisms), a model to help understand

women’s participation in the DCI. Agent-driven mechanisms provide an original,

empirically grounded, and theoretically driven explanation of why women participate in

the DCI.

Further theoretical development lead to the proposal of an emerging theory titled

the Acts of Agency theory (proposed in Section 6.2.4). This theory further extended and

integrated the previous models, including the Sphere of Influence (SoI), Events of

Interaction (EoI) and Five Acts of Agency (FAA).

Chapter 6 also discussed the value of using the Five Acts of Agency model in an

applied manner. Existing strategies and initiatives aiming to foster women’s participation

can be developed from, or evaluated by, using the Five Acts of Agency model. Eleven

recommendations are made, with each recommendation providing evidence of how

agent-driven mechanisms can foster women’s participation in the DCI.

Therefore, the two research questions, posed at the conclusion of Chapter 2–

‘What are the influences on women’s participation in the DCI?’, and ‘How can these

influences be understood?’–were addressed by developing and then applying a series of

three frameworks to the analysis of empirical data. This process has resulted in an

understanding of the influences on women’s participation in the DCI through the

development of three new models: (1) the Sphere of Influence (SoI), (2) the Events of

Interaction (EoI), and 3) the Five Acts of Agency (FAA), and an emerging theory titled

the Acts of Agency theory. The remainder of Chapter 7 summarises how the research

addressed the research problem and the research questions.

Section 7.1, explains how the research contributes to systematically addressing

the two research questions and the research problem. Section 7.3 summarises the

empirical, theoretical, methodological and applied contributions. There are limitations to

the findings and conclusions, and the need for future research is not surprising as the

nature of exploratory research is to reveal more questions than it answers. As Section 7.4

concludes, these limitations, in many instances, inform the direction for future research.

7.1 ADDRESSING THE RESEARCH PROBLEM

The research problem addressed by this research was ‘Why do women participate

in the Australian Digital Content Industry (DCI)?’ Two key research questions addressed

this research problem. Question 1 asked, ‘What are the influences on women’s

participation in the DCI?’, and Question 2 asked ‘How can these influences be

understood?’

Employing an exploratory case study approach, the scope of the investigation

focused on women’s participation within games and multimedia organisations in

Brisbane (Queensland, Australia). The city of Brisbane was, at the time of the

investigation, the second largest digital games hub in Australia (as noted in Section

4.1.1). The primary empirical data collected for the case study included semi-structured

interviews with 18 women (employed as interactive content creators in 11 different

games and multimedia organisations), and three industry stakeholders (including a male

DCI worker and a male employer). Secondary data included over 60 individual samples

of data, such as excerpts from websites and industry reports relevant to the DCI.

Underpinned by a Critical Realist ontology, the case study aimed to provide empirical,

theoretical, and explanatory insights into the research problem.

7.1.1 THE INFLUENCES ON WOMEN’S PARTICIPATION

The findings support previous research that suggests women are under-

represented in the DCI (as reported in the findings in Chapter 5). Analysis of the data

using the three frameworks proposed in Chapter 3 reveals a range of influences on

women’s participation. The influences include those previously reported in both the DCI

and ICT industry-related literature; influences such as other people (for example,

parents), organisational cultural practices, media stereotypes and access to resources. In

addition, the findings illuminate influences not noted in previous research, such as the

type of clothing worn by DCI workers, which can signify gendered practices. The

findings identify a wide range of social, cultural, mediated and resource influences that

influence women’s participation in the DCI context.

Many of these influences have been identified in the ICT literature. However, the

findings add a deeper insight into the influences. Similar to previous research in the ICT

literature, for example, the findings reveal that women in the DCI do face a workplace

culture that fosters working long work hours. However, findings also reveal that the

influences resulting from the interaction between the environment and person are

complex and emergent. Although women in the DCI face expectations to work long

hours, the actual influence of work hours differed. Not all interactive content creator

roles entailed long working hours; the practice varied between DCI organisations, and

the individual women themselves played a role in choosing to work long hours (See

these findings in Chapter 5). Thus, in identifying what influences exist the research has

also highlighted that there is need for a richer understanding of those influences. Care

should be taken in making a blanket statement that ‘long hours’ are an influence. There is

a range of influences on women’s participation and these vary with regards to the

environment and person.

This empirical data provides a valuable contribution, as there is little previous

empirical research into women’s participation in the DCI. However, to address the need

(recognised in Chapter 2) for unified rather than fragmented insights, the research

approach was further refined to identify influences, and to understand these influences in

a unified manner.

7.1.2 THREE FRAMEWORKS WITH WHICH TO IDENTIFY AND UNDERSTAND INFLUENCES

The application of three different frameworks to the analysis of multiple sources

of empirical data resulted in the identification of existing influences and the further

understanding of these influences. As Chapter 3 proposed, these three frameworks were:

1) analytical, 2) theoretical, and 3) ontological. The sequential application of each of the

three different frameworks, to the analysis of empirical data, provided valuable insights

into the influences on participation. Each framework revealed different influences and

offered a way in which to understand these influences. The development and

application of these three frameworks (analytical, theoretical, and ontological) is an

original approach to identifying and understanding the influences on women’s

participation in the DCI, and is a major contribution of this research.

FRAMEWORK 1 (THE ANALYTICAL FRAMEWORK ): E, P AND I

The first framework (analytical) was developed from a review of existing

research (Section 3.1). The analytical framework fostered an investigation into three

categories of influences: 1) Environment (E), 2) Person (P), and 3) the Interaction

between E and P (I). Chapter 3 argued that there was a need to investigate two key

categories of influences, the environment and the person. Previous research proposed that

various sets of environmental factors and person characteristics could influence women’s

participation in either the ICT industry or the DCI. Analysis of the literature also

suggested that investigating the interaction between the environment and person

characteristics could be useful in understanding influences in a more holistic and

integrated manner–with some researchers even suggesting that a focus on interaction was

vital. Although approaches such as Trauth’s Individual Differences Theory of Gender

and IT (IDT) (Trauth et al., 2004) recommended that the interaction between influences

be considered, little research directly focused on the interaction. Therefore, the first

framework was developed in order to achieve this analysis.

Accordingly, the empirical data collected for the exploratory case study was

analysed using the first framework. Through a process of pattern analysis, the empirical

data was categorised into E, P or I categories. As presented in Section 5.2, this analysis

provided a detailed granularity of the three categories in Framework 1. The Environment

(E) was found to be comprised of four dimensions: 1) social, 2) cultural, 3) mediated,

and 4) resource, with each dimension having several specific properties. The social

dimension (1) included the properties of social agent (including parent, peer, and

colleague) and social context (home, school, and work); the cultural dimension (2)

included properties such as economy and language; the mediated dimension (3) included

the property of medium; and the resource dimension (4) included properties such as

technology. Specific influences identified in the social dimension included the ratio of

male to female colleagues in the workplace. The mediated dimension included the

stereotypes conveyed by different mediums, including television. The resources

dimension included access to technology, including computers, which could foster skill

development.

The Person (P) category had only one dimension, social identity, which had two

major properties, 1) occupational and 2) gender-related, and two minor properties, age

and ethnicity. As discussed in Section 6.1, the ‘social identity’ concept in the Person (P)

category helps to describe the person involved by recognising that, regardless of an

individual’s qualities and capacities, one is initially judged by society’s social

stereotypes, which can include gender stereotypes. The stereotypical views that an

individual or other people have may influence or ‘filter’ the types of interactions a person

experiences. Thus, a person’s social identity forms a type of a filter of experiences.

A woman faces gender stereotypes, such as their not being passionate about, or

adept with, technology. Compounding the influence of gender stereotypes were the

occupational stereotypes suggesting that DCI workers possessed a natural technical

prowess. Thus, there were tensions between the two major properties, gender and

occupation, within the social identity: Gender stereotypes suggest women are not

technical, and occupational stereotypes suggest DCI workers require technical skills. As

was discussed in Section 6.1, the two refined categories of E and P, and their dimensions

and properties, are expressed as the Sphere of Influence (SoI) model.

Inductive analysis of the data for the third category in Framework 1, Interaction

(I) between environment and person, proved more challenging because of its dynamic

nature. As Section 5.2.3 explained, the research approach drew on the underpinning

ontology, or more specifically on Bhaskar’s emphasis on recognising events (that

individuals may or may not be aware of), to move beyond the analytical impasse. The

empirical data supports this approach as participants’ accounts of their participation were

often couched in terms of events. The refinement of the Interaction (I) category led to the

identification of 10 events of interaction, which include gaining entry into the industry

(‘Getting In’), negotiating salaries (‘The Catch 22’), and becoming a parent

(‘Motherhood’); these are fully described in Section 5.2.3. As discussed in Section 6.1.4,

these 10 events informed the development of a second model resulting from the analysis

of empirical data using Framework 1: the Events of Interaction (EoI).

Thus, the key contributions resulting from the analysis of the empirical data,

using Framework 1, were the development of two conceptual models: 1) the Sphere of

Influence (SoI), which illuminates the environment and person characteristics that may

influence women; and 2) the Events of Interaction (EoI), which illuminates particular

events over a person’s lifetime. The SoI provides descriptions of the dimensions and

properties of the Environment (E) and Person (P) characteristics that may influence

women’s participation; this, in turn, raised awareness of the nature of the Interaction (I)

between these two categories. The EoI model extends the description fostered by the SoI

by highlighting key events of Interaction (I) between the E (Environment) and Person

(P). Together, the models begin to offer a unified approach to identifying the influences

on women’s participation that manifest in the interactions between the environment and

the person.

The analysis of empirical data using the analytical framework, and the two

resultant models, helps provide a richer understanding of why women participate in the

DCI. There were two key strengths of employing Framework 1. The analytical

framework fostered the development of models to organise insights, and then provided a

body of existing research with which to compare findings. As discussed in Chapter 6, the

analysis of empirical data using the analytical framework helped reveal many influences

that previous research had identified, and this strengthens the likelihood that those

influences are involved in women’s participation.

The two models emerging from Framework 1 also provide a unified approach to

understanding the influences on women’s participation. There are four strengths of the

two models: 1) they help avoid essentialist propositions about gender, 2) they allow for

linkages between macro and micro-level influences, 3) they provide insights that are

complementary to the underpinning ontology; and 4) they allow for influences to be

considered with regards to their temporality. First, in offering a description of the Person

(P) category, an overriding concern of this research has been to avoid any essentialist

propositions regarding gender. Indeed, the choice of the term ‘person’ offers a non-

gendered label. Second, as discussed in Section 6.1, the models encourage analytical

linkages, which can offer a deeper level of insight or a more unified understanding of

influences. The Environment (E) can be considered across different levels, both the

macro (for example, global or national) and micro level (for example, local setting such

as organisation), thus highlighting links between influences such as global economic

trends (macro) and the local hiring practices (micro). Third, both models provide

insights into Bhaskar’s ‘empirical’ and ‘actual’ domains of reality by illuminating both

the experiences of individuals and the events that they may or may not be aware of.

Fourth, both models begin to recognise that influences manifest within a temporal

context. The SoI proposes that certain social agents within the social dimension of the

environment, such as parents, may exert a stronger influence on women’s participation at

certain stages in their lives (for example, in childhood, as discussed in Section 6.1). The

EoI recognises that events, such as gaining skills, occur over a lifetime, from early

childhood through to adulthood when a person is in the workforce. These models help

form a richer view and foster a holistic understanding of the influences on women’s

participation.

The evidence presented and discussed in Chapters 5 and 6, and summarised in the

section above, shows that the Sphere of Influence (SoI) and the Events of Interaction

(EoI) models are key contributions of this research. Their development and application

has enabled a richer understanding of the influences on women’s participation. The

evaluation of their contribution shows that the use of the two interrelated analytical

models has been useful in providing that understanding. However, both these models

primarily facilitate description of the influences. While rich descriptive insights are a

valuable contribution, there are limitations to description alone. Thus, there were benefits

in applying the second framework, the Human Agency Multi-Theory Scaffold (HAMTS)

and the third framework, Bhaskar’s three domains of reality (or the ontology of Critical

Realism), as both facilitate further theorisation and explanation (rather than description)

of the influences.

FRAMEWORK 2 (THE THEORETICAL FRAMEWORK ): HUMAN AGENCY MULTI -THEORY SCAFFOLD (HAMTS)

The second framework was a theoretical framework comprised of concepts from

four Human Agency theories (as proposed in Section 3.2). Chapter 3 argued that agency

theory offered a suitable approach with which to investigate the interactions between the

environment and the person. Chapter 3 also emphasised the methodological implications

of choosing which theory, or theories, to use as a scaffold. In response, the concept of

the Multi-Theory Scaffold (MTS) was proposed. The MTS is a methodological tool that

guides the use of more than one theory as a scaffold. The strength of the MTS is that it

encourages more than one theory to be used, as different types of theories (such as meta,

critical, middle-range and operationalised theory) foster different types of understanding.

Although the MTS could employ various theories, in this research, the MTS was

populated with Human Agency theories, and this Human Agency Multi-Theory Scaffold

(HAMTS) forms the basis of Framework 2. As identified in Section 3.2.4, Framework 2

is comprised of 11 concepts selected from the following four Human Agency theories: 1)

the meta theory of Giddens’ Structuration Theory (ST); 2) the critical theory of Connell’s

Social Theory of Gender (STG); 3) the middle-range theory of Bandura’s Social

Cognitive Theory (SCT); and 4) the operationalised theory of Lent et al.’s Social

Cognitive Career Theory (SCCT). The MTS is a further key theoretical contribution

emerging from this investigation.

analysed using the second framework. Through a process of pattern analysis, the

empirical data was categorised into 11 categories (the agency theory concepts), including

norms, power, transgression, and self-efficacy. As presented in Section 5.2, the

examination of the data using the HAMTS revealed that there was empirical evidence to

support each of its concepts. Thus, Human Agency theory offered a useful theoretical

perspective with which to analyse empirical data to help reveal both a diverse range of

influences, and to offer a way to understand women’s participation.

The concepts in the HAMTS also challenged the empirical data, and the tensions

between existing theory and the empirical data strengthened the development of original

theoretical insights. For example, although ST (Giddens, 1984) and SGT (Connell, 1987)

emphasise the role of societal power, there were few examples of this type of power in

the empirical data. In contrast, the participants’ comments were interpreted as

emphasising individual power, which ultimately led to a focus on the causal capacity of

the individual. Thus, analysis of the empirical data, using agency theories, began to

emphasise the role a person played in regards to their participation. This theoretical

perspective proved central when synthesising the findings.

Previous research into women’s participation did not draw heavily on Human

Agency theory, even though previous research in the ICT literature alluded to the

relevance of this theory (as Section 3.2.3 noted). Therefore, this investigation has

provided a new and fruitful theoretical base from which to understand the phenomenon

of women’s participation in the DCI. Specifically, two key conclusions emerge from the

application of Framework 2. The first conclusion relates to the value of a MTS to the

research process, and the second relates to the value of agency theory in addressing the

research problem. As discussed in Section 6.1.6, the MTS made three primary

contributions to the research process: 1) sensitised the researcher to the phenomenon

under investigation by offering a wider range of concepts for consideration; 2) assisted in

the analysis of empirical data; and 3) fostered theorising about the phenomenon. The

MTS fostered different levels of abstraction and strengthened the theoretical triangulation

between data and theory, which led to greater levels of confidence in the findings and the

research outcomes. Second, agency theories (within the HAMTS) offered a range of

concepts that drew attention to influences, such as power, not readily identified from

empirical data. In addition, agency theories fostered further understanding of influences.

For example, where Framework 1 identified the role of resources, agency theories

explained why these resources could influence participation. The value of employing

more than one agency theory was that different theories offered different explanations of

influences. Where Giddens emphasises the role of power when accessing resources,

Bandura emphasised that access to resources plays a role in a person’s self-efficacy. The

use of the HAMTS, therefore, addresses the call in the literature for a theoretical

understanding of the phenomenon of participation by facilitating an empirically

grounded explanation of the phenomenon.

Chapter 3 foreshadowed many of the benefits that emerged from the employment

of the HAMTS in the research. An unexpected benefit was the value of using Human

Agency theory to abstract the underlying mechanisms. That is, the concepts from

Framework 2 helped reveal the mechanisms that Framework 3 aimed to identify. The

complementary nature of Frameworks 2 and 3 is summarised shortly.

FRAMEWORK 3 (THE ONTOLOGICAL FRAMEWORK ): BHASKAR ’S THREE DOMAINS OF REALITY

The third framework was the ontological framework, comprised of three

concepts, the ‘three domains of reality’ proposed by the Critical Realist Bhaskar (1978,

1989): ‘empirical, actual and real’ (See Section 3.2). Chapter 3 argued that there was a

need for a multi-paradigmatic, deeper understanding of influences. Critical Realism’s

emphasis on the abstraction of mechanisms underlying people’s experiences appeared to

offer a suitable philosophical or ontological perspective. In addition to specifically

employing Bhaskar’s concepts in Framework 3, the research process itself was

underpinned by a Critical Realist ontology.

analysed using the third framework. Through a process of pattern analysis, the empirical

data was categorised into three types: 1) empirical, 2) actual, and 3) real. As presented in

Section 5.2, this analysis revealed mechanisms underlying the interactions between the

environment and person. Although Bhaskar’s ‘empirical’ domain encouraged a

description of the entities such as environment and person, earlier data analysis revealed

that many of the entities in the empirical domain had been previously identified when

using Framework 1 (as captured by the SoI model); consequently, there was no need to

investigate the category further. Similarly, the consideration of Bhaskar’s actual domain

encouraged a description of events of which participants may or may not be aware. Data

analysis using Framework 1 had also revealed the events (as captured by the EoI model),

and no further analysis was conducted.

Investigating Bhaskar’s real domain entailed the abstraction of mechanisms

underpinning interactions; this (as Section 6.1.8 noted) proved more challenging. The

abstraction was addressed using a number of analytical approaches. A key approach

entailed following Danermark et al.’s (2002) recommendation that existing theory

(utilising an abductive logic) be used. Although the HAMTS had not explicitly considered

mechanisms, the agency theories (in Framework 2) drew attention to the causal capacity

of individuals. This insight and further analysis of the empirical data (utilising

retroductive logic) led to the abstraction of mechanisms in general, and of agent-driven

mechanisms in particular. Agent-driven mechanisms are driven by the individual, in

contrast to those that arise from the environment. Therefore, the influences on women’s

participation include agent-driven mechanisms, as the following section will further

clarify. Therefore, Critical Realism and, more specifically, Bhaskar’s ‘domains of

reality’ offered a useful approach to describing the phenomenon of women’s

participation by identifying influences and in understanding these influences; it also

offered a useful approach (based on causal tendencies) to explaining why women

participate in the DCI.

THE AGENT -DRIVEN MECHANISMS

As noted, each of the three frameworks revealed different types of influences;

this was because each individual framework offered a different perspective with which to

analyse the empirical data and understand the influences on women’s participation.

Although each framework illuminated different influences, there were common findings

across all three–each investigated the interaction between the person and their

environment, and each revealed that the individual plays an important role in their

participation.

The first framework, informed by existing research, recognised that it is integral

to consider the interaction between the environment and person. Analysis of empirical

data using Framework 1 revealed that both the characteristics of the environment and

person play a role in participation.

The second framework, informed by existing agency theory, inherently offered

several specific concepts (such as self-efficacy) to explain interaction between the person

and their environment. Analysis of empirical data using Framework 2 revealed how the

person (or the agent) has a potentiality for causal action.

The third framework, informed by a Critical Realist ontology, offered three

concepts, with an emphasis on mechanisms that underpin the interaction between the

person and their environment. Analysis of empirical data using Framework 3 revealed

that mechanisms include those that arise from the causal effect of the person.

Through the conduct of the research, it became apparent that, although each

framework helped to identify certain types of influences, it was the synthesis of the

insights from the application of each framework which was most helpful in addressing

the research problem. This synthesis led to the original conceptualisation of agent-driven

mechanisms.

The key, original contribution from the synthesis of insights from the application

of the three frameworks was the conceptualisation of agent-driven mechanisms. Agent-

driven mechanisms provide an original means of explaining women’s participation. The

conceptualisation of agent-driven mechanisms is encapsulated in the development of

both a model and an emerging theory. The conceptual model–the Five Acts of Agency

(FAA) – provides a relational, conceptual model, which identifies 10 agent-driven

mechanisms that underlie the emergent interactions between the person and their

environment. The 10 agent-driven mechanisms are: 1) accessing, 2) imagining, 3) doing,

4) belonging, 5) sharing, 6) problem-solving, 7) transforming, 8) emotion, 9) ethical, and

10) being (See Section 6.3.3). The agent-driven mechanisms conceptualise the causal

effect an individual may have when interacting with their environment. For example,

individuals can harness agent-driven mechanisms (such as collaborating and problem

solving) to foster their participation.

Although women face constraints and supports (from both their environment and

themselves), they also play an integral role by having a casual effect on their

participation. Participation is a result of an individual’s actions (or inaction) in response

to the environmental characteristics and their personal characteristics. Where previous

research focused on the overwhelming circumstances women face in regards to their

participation, the concept of agent-driven mechanisms recognises that both the

constraints and supports are, to some extent, in the hands of the women themselves.

Women are not positioned as hapless victims; they can, and do, have the capacity to have

a causal effect–to challenge the circumstances in which they find themselves. Thus, the

agent-driven mechanisms, as conceptualised in the Five Acts of Agency (FAA), can

influence women’s participation in the DCI. The conceptualisation of the 10 agent-

driven mechanisms, in the Five Acts of Agency (FAA) model, is the key contribution

emerging from this research.

Further theoretical advancement of the concept of agent-driven mechanisms

involved the proposal of an emerging theory, the Acts of Agency theory. The theory

provides further explanation of women’s participation in the DCI by integrating the three

models that emerged from the findings: 1) SoI, 2) EoI, and 3) FAA. Used sequentially,

each model reveals different influences, moving from a surface description of

characteristics through to a deeper account of underlying mechanisms. The SoI identifies

the characteristics of the environment and person that may influence participation. The

EoI identifies events (where the environment and person interact) that may influence

participation. The FAA identifies the mechanisms (which manifest in the interaction) that

a person can employ to influence their participation. Additionally, the theory extends the

three models by introducing the elements of the agent’s actions, reflection and

temporality (as discussed in Section 6.2.4). This theorisation draws heavily on both

Human Agency theory and Critical Realism literature. The central premise of the Acts of

Agency theory, then, is that although participation involves both the environment and

person, the causal effect of the individual is central.

There are two major benefits of using agent-driven mechanisms to understand

women’s participation: 1) a causal explanation and, 2) a non-essentialist perspective.

The first benefit is that they move findings beyond a description of a phenomenon

towards an explanation of women’s participation. Describing the influences on

participation is important, and this research has achieved that by analysing the empirical

data and proposing models such as the SoI and EoI. However, this research also found

that the interactions between the influences in the environment and the person manifest

in a mutable and dynamic manner, which inhibits a definitive description of their

influence on participation. Agent-driven mechanisms are able to reduce this complexity

by recognising the causal tendencies. In the context of this research, the Acts of Agency

theory suggests that even though individual women may have different experiences,

potentially all may draw on this set of agent-driven mechanisms to foster their

participation.

The following example serves to illustrate how agent-driven mechanisms foster a

causal explanation of the phenomenon of women’s participation. Framework 1 found

that resources, such as computers, are part of the group of environmental influences

(resources are one of the four dimensions of the SoI model) that foster participation. The

analysis also found that these resources assist women to develop skills and self-efficacy

and, ultimately, to challenge the gender stereotypes that they experience (as illustrated by

the event ‘Learning Recipes’ in the EoI model, and further explained by drawing on the

agency theories in Framework 2). The SoI and EoI models facilitated a descriptive

account of the different types of resources (for example, computer models and software

packages) within the environment category, and a range of individuals within the person

category (for example, different ages, personal backgrounds, and roles as interactive

content creators). Agent-driven mechanisms (in the Five Acts of Agency model) provide

a further explanation of the influences of resources by acknowledging the causal effect of

the individual. For example, when a woman (person) uses a resource, such as software,

the mechanism of ‘doing’ (that is acquiring and practising skills) occurs. The agent-

driven mechanism of ‘doing’ occurs regardless of the type of resource (which can

change) or characteristics of particular individuals (which can also vary). Thus, agent-

driven mechanisms reveal underlying causal mechanisms, rather than surface accounts

The second benefit of agent-driven mechanisms is that they offer a non-

essentialist approach to understanding the causal effect of individuals. Although an

individual characteristic such as gender is acknowledged in the SoI (and, consequently,

as a component of the Acts of Agency theory), gender itself is not viewed as the only

reason a person may or may not participate. The characteristic of gender gives rise to

other influences, including the stereotypes a person faces which might constrain their

capacity to feel as though they belong. However, in response, a person can find ways

(through their actions) to overcome these constraints. For example, an individual

recognising that they face exclusion in the workplace due to their gender may actively

seek access to resources that will minimise this exclusion. Following the Five Acts of

Agency model, they would (in this instance) be employing the mechanism of ‘accessing’.

Thus, although women may face limited access to resources, their participation to

some extent lies in their own hands as individuals. However, women are not to be held

entirely accountable for their participation. They are not responsible for the gender

stereotypes they face, or the ‘old boys’ networks that impede their access to resources.

Although the Acts of Agency model acknowledges that participation is within the control

of individual agents, women’s participation should always be considered with reference

to the characteristics of the environment.

COMPLEMENTARY FRAMEWORKS

The synthesis of insights from the data analysis, which lead to the original

conceptualisation of agent-driven mechanisms, arose from the complementary aspects of

the three frameworks. All three frameworks focused the analysis of the empirical data on

the interaction between the environment and person (as Chapter 3 detailed). Further

complementary aspects of the three frameworks emerged during the research process.

First, there was a complementary focus between the analytical framework (Framework 1)

and the first two domains of reality proposed by the philosophical framework

(Framework 3). Framework 3 entailed the analysis of empirical data using the concepts

of the three domains of reality: 1) empirical, 2) actual, and 3) real. The empirical domain

seeks to identify the experiences of people, and the actual domain seeks to identify

events of which people may or may not be aware. The SoI, a model emerging from the

refinement of Framework 1, provided an insight into the empirical domain, as the SoI

entailed the interpretation of participants’ accounts of their experiences. The EoI model

provided an insight into the actual domain of reality, as the EoI fosters a description of

events that participants experience, but of which they may not be aware. The findings

from the analysis of empirical data utilising the analytical framework (Framework 1)

were of assistance when analysing data using the ontological framework (Framework 3).

Importantly, this complementary aspect links previous research (recognised by

Framework 1) and an ontological stance (Framework 3).

Second, there was a complementary focus between the theoretical framework

(Framework 2) and the third domain of reality, the real domain, in Framework 3. The

real domain seeks to identify the mechanisms underpinning the experiences of people.

Accordingly, Framework 3 entailed the analysis of empirical data to allow for the

abstraction of the mechanisms underlying the interactions between the environment and

person. As discussed in Chapter 6, although the ultimate aim of a Critical Realist

ontology is the abstraction of mechanisms, in practice, there are challenges in this

process (See Section 6.2.6). Following recommendations by Danermark et al. (2002),

different forms of logic were employed; these included an abductive logic when drawing

on theory to abstract mechanisms from the empirical data, and a retroductive logic when

abstracting mechanisms from the empirical data. The investigation found that both the

empirical data and the Human Agency theories from Framework 2 assisted in the

abstraction of mechanisms. However, Human Agency theories proved to be particularly

useful in the abstraction of mechanisms as they illuminated the causal capacity of the

individual. Although Human Agency theories do not specifically offer causal

mechanisms (in a Critical Realist sense), the theories sensitised the researcher to the

underlying causal mechanisms at play in the interaction between Environment (E) and

Person (P) through concepts such as structuration (Giddens, 1984), reciprocal

relationships (Bandura, 1986), and causal pathways (Lent et al., 1994). Thus, there was a

complementary aspect between the theoretical framework (Framework 2) and the

ontological framework (Framework 3). The findings from the analysis of empirical data

utilising the theoretical framework (Framework 2) were of assistance when analysing

data using the ontological framework (Framework 3). These conclusions suggest that not

only are there parallels between Human Agency theory and Critical Realism in a

theoretical sense, but that the different perspectives also offer a complementary approach

in practice.

In addition, the models developed from the application of the three frameworks

appear to align with Danermark et al.’s (2002) six-stage model of explanation. Both the

SoI and EoI provide insights into Stages 1 and 2 of Danermark et al.’s (2002) model,

because they foster a description of the entities involved in the phenomenon, help

organise insights, and provide a resolution of data. As discussed in Chapter 6, the MTS is

beneficial in Stage 3 of Danermark et al.’s (2002) model. Where Danermark et al.,

(2002) emphasised the role of existing theory, the MTS provides a guide to the use of

multiple theories. The strength of the MTS is that more than one theory can foster the

abstraction of a wider variety of mechanisms. In this way, the MTS could well be

expanding the work of Danermark et al. (2002). The FAA model and Acts of Agency

theory align to Stages 5 and 6 of Danermark et al.’s model, where the emerging

conceptualisation of agent-driven mechanisms is supported by both empirical data and

further theoretical work.

The use of the three frameworks in an integrated manner provided a valuable

approach through which to understand the influences on women’s participation and,

ultimately, to understand why women participate in the DCI. The application of an

analytical framework built on existing research (Framework 1), facilitated the

development of two of the new models that arose from this research: the SoI and the EoI.

The SoI captured the environment and person characteristics, and the EoI illuminated

events where the environment and person interact. However, the application of

Framework 1 also confirmed the need for a different approach with which to better

understand the interaction between influences, hence supporting the value of the

following two frameworks. The application of the Human Agency Multi-Theory Scaffold

HAMTS (Framework 2) confirmed the value of using theory to inform the research

process, and specifically confirmed the value of using multiple Human Agency theories

to understand the influences on women’s participation. The value of using Critical

Realism as an ontological framework (Framework 3) lay in the abstraction of

mechanisms, which provided a form of explanation of women’s participation.

Individually, each of these frameworks offered a different approach to understanding the

influences on women’s participation; however, together, they offered a deeper, unified

understanding of the research problem.

7.2 CONTRIBUTION

Addressing the research problem and research questions resulted in four principle

forms of contribution: 1) empirical, 2) theoretical, 3) methodological, and 4) applied.

Empirical contributions include a volume of rich data, mostly sourced from the accounts

of women who work in the DCI. The SoI and EoI models helped to organise this

empirical data. Theoretical contributions include the original conceptualisation of agent-

driven mechanisms, encapsulated in the FAA model and in the emerging theory of Acts of

Agency. The methodological contribution includes the development and application of

the Multi-Theory Scaffold (MTS) –an analytical tool for choosing different types of

existing theories when using a theoretical scaffold. The applied contribution includes the

apparent usefulness of the proposed agent-driven mechanisms as a way to plan and

evaluate strategies supporting women’s participation.

1) EMPIRICAL CONTRIBUTION

This research provided an empirical contribution through the conduct of an

exploratory case study. The case study entailed collection of multiple sources of

empirical data, including semi-structured interviews with women working in the DCI and

stakeholders from the industry. Secondary data included excerpts from websites, the

media, and industry documents (as summarised in Appendix 17). Analysis of the

empirical data resulted in rich descriptive insights into the phenomenon, including

characteristics of the context (Australian DCI) and the population involved (women

working in the DCI). Insights reveal a range of influences on women’s participation such

as long work hours, gender and occupational stereotypes, and the availability of

resources. Such descriptive insights are particularly valuable as there has been little prior

empirical research on the participation of women in the specific context of the Australian

DCI. To the author’s knowledge, this is the only dissertation and only the second

academic study in the last decade [Roan and Whitehouse’s (2007) study was the first] to

explore women's participation in the Australian DCI.

2) THEORETICAL CONTRIBUTION

This research provided a theoretical contribution through the synthesis of empirical

data, concepts from previous research, the perspective of Human Agency theory and a

Critical Realist ontology. This synthesis has led to the development of four empirically

derived, theoretically driven models: 1) The Sphere of Influence (SoI); 2) The Events of

Interaction (EoI); 3) The Five Acts of Agency (FAA); and 4) the emerging theory of Acts

of Agency. The SoI, EoI, and the FAA support the emerging theory of Acts of Agency,

which in turn, provides a starting point for further theorisation.

3) METHODOLOGICAL CONTRIBUTION

This research offers a methodological contribution in two ways: first, the

investigation provides an exemplar of an empirical, exploratory case study underpinned

by Critical Realist ontology (of which, as Chapter 4 noted, there are few); second, the

investigation presents an original approach to employing multiple theories in a

theoretical scaffold. With respect to the first contribution, the study illustrates the

argument for using both a case study approach and the ontology of Critical Realism: both

value multiple sources of evidence and both foster empirical, theoretical and explanatory

insights.

With regard to an empirical exploratory case study employing a theoretical

scaffold (as outlined in Chapter 3), there are contentions surrounding the application of

existing theory in a largely inductive case study. Embracing existing theory as a part of

the research process can lead to questions about how original insights emerge from a

study, or may suggest a repudiation of empirical realism. However, this study has argued

that a theoretical scaffold serves as a tool with which to understand the empirical data. It

does not superimpose theory; rather, it stimulates theoretical insights from the empirical

data. This investigation clearly articulated the use of existing theory as a scaffold and has

developed and applied a Multi-Theory Scaffold (MTS), a framework that embraces the

strengths of using multiple theories to scaffold research by asking the researcher to

consider the types of abstraction that different theories offer a study. The MTS is a

valuable research tool for three key reasons; it can: 1) sensitise the researcher, 2) assist in

analysis of empirical data, and 3) assist in theorising the phenomenon. The MTS,

therefore, can be of value to researchers who are interested in using existing theory to

scaffold research.

4) APPLIED

In addition to its academic merits, the research offers an applied contribution.

Section 2.3 identified the need for theoretically informed interventions to foster the

recruitment and retention of women, and to address issues of social inclusion of women

as a minority group in society. The findings reveal that agent-driven mechanisms (in the

FAA model) offer an approach to planning and evaluating strategies and interventions (as

discussed in Section 6.3). For example, both the women in the current case study and

previous research in the ICT domain have identified the need for role models. Women’s

access to suitable role models can foster their capacity to visualise or imagine their

participation in industry and career pathways. In other words, the strategy of offering role

models supports the individual’s capacity for the agent-driven mechanism of ‘imagining’

(categorised in the ‘Connected Act’ within the Five Acts of Agency model); therefore, the

agent-driven mechanisms help to explain why certain strategies can support women’s

participation.

7.3 LIMITATIONS OF THE RESEARCH

As with all research, findings must be interpreted with regard to their limitations.

Several limitations have been noted and are now addressed.

One limitation is that in aiming for a non-essentialist approach to researching

women’s experiences, the investigation has distanced itself from an explicit focus on

gender. However, gender is a central concern as evident by the choice of sample

population (women) and the inclusion of a critical theory in the MTS, which has

introduced the axiological considerations often associated with feminist literature.

However, rather than being a limitation, this distancing from gender is the strength of the

resultant models because they foster a non-essentialist approach. The Sphere of Influence

(SoI) and Five Acts of Agency (FAA) models, therefore, can be transferred to research

concerned with the participation of other social identities (for example, ethnicity and

A further limitation is that the research approach requires a high level of skill on

the part of the researcher. Critical Realism is viewed as an approach that is not suitable

for a novice researcher (as noted in Pawson, Greenhalgh, Harvey & Walshe, 2004);

rather, it is one that requires experience, both in research and in subject matter (Dobson

et al., 2007). In response to this limitation, every effort was made to make the logic,

methodological approach, and high level of researcher reflexivity apparent.

Consequently, it is hoped that the considered choices that the researcher has made are

evident in the investigation.

A further limitation of the research is that findings report a diverse and complex

range of influences, and an in-depth discussion of all of these influences was beyond the

scope of this dissertation. Such constraints resulted in surface, rather than deep accounts

of certain influences. Further consideration of how the characteristics of the environment,

such as work practices (for example, work hours and the typical clothing of DCI

workers), and physical features of the environment (for example, open space layout of

workplaces) is required. A similar limitation of the applied outcomes is that only a few

strategies were discussed (in Chapter 6). In response to this limitation, there is a need for

further research which reduces the depth or breadth of enquiry, or which focuses on

specific influences.

A further limitation lies in identifying the influences pertinent to the DCI context.

Many of the influences that this study identified have surfaced in previous research in the

ICT related literature. Nuances were noted in the discussion; however, further

consideration of the DCI specific influences is required.

There are also limitations to the models emerging from the application of the

frameworks. The Sphere of Influence (SoI), Events of Interaction (EoI), Five Acts of

Agency(FAA), and emerging theory of Acts of Agency are what Layder (1998, p.78)

described as “interim products”. In other words, there remains the possibility for revision

of these models and theory in the light of further evidence. For example, further research

may reveal other dimensions and properties of the SoI model. Similarly, the ten events

identified in the EoI model are relevant to the particular sample of women interviewed

for this investigation; further interviews with other DCI women could reveal other

events. Thus, this investigation presents inferences rather than definitive conclusions, and

alternative interpretations are always possible.

Similarly, any conclusion regarding the explanatory strength of Human Agency

theories must consider the methodological concerns regarding the use of existing theory.

This includes acknowledging that no one theory can provide an ultimate explanation of

such a dynamic and complex phenomenon as participation. In response to this limitation,

there is the need to reiterate that the strength of a theoretical tool such as the MTS is that

it does not recommend a specific theory; rather, it asks the researcher to consider which

theories, based on their level of abstraction, are useful to the research problem at hand.

This would not necessarily entail using agency theory; a researcher can use any theory in

the MTS that they view as relevant. The MTS simply asks the researcher to recognise that

different types of theories may lead to different insights. Even though agency theories

appear relevant to the research problem, other theories may be just as fruitful.

7.4 FUTURE RESEARCH OPPORTUNITIES

The limitations outlined above offer opportunities for further research. Areas of

future research, or what Layder (1998) might describe as “offshoots”, include the in-

depth research of specific influences (such as culture), and further development of the

models. As noted earlier, the need for future research is not surprising, as the nature of

exploratory research is to reveal more questions than it answers. Table 54 summarises

future areas of research noted throughout the dissertation; these are as follows.

The research provides a valuable insight into a wide range of influences. Future

research could refine these substantive insights, including greater in-depth consideration

of specific influences identified from the empirical data, such as work hours. There is

also a need for further expansion of specific concepts, including identity filter, gender

triggers, and temporality.

The research will be of value to other researchers investigating participation.

Although the Acts of Agency is a substantive theory developed from the study of one area

of investigation (participation) within a specific context (Australian DCI) and population

(female interactive content creators), there is scope for application of the emerging

theory (and the associated models) to other research problems and domains where the

aim is to understand the phenomenon of participation. As the models recognise social

identity, rather than fixed traits such as gender, they should be useful in research that

considers participation by other identities (such as male, older workers), and within other

contexts and industries.

The research provides several models, which individually and together offer a

way to understand women’s participation in the DCI. Future research could employ the

models as sensitising devices for future research. Both the SoI and EoI, as theoretical

tools, can sensitise the researcher to environment, person and interaction characteristics

that may influence women’s participation. Similarly, the FAA can serve as a sensitising

device for future research investigating mechanisms. This may entail the FAA model

forming an explicit analytical framework for further collection and analysis of empirical

The models provide a strong basis for theorisation. Future research could extend

the theoretical tools emerging from this research through further empirical and

theoretical research.

Table 54 Examples of future areas of research resulting from this dissertation Future area of research

Examples as noted in the dissertation

Further empirical work to investigate specific influences

-Future research could use a sampling strategy where the differences between the games and multimedia organisations can be further investigated -Future research could focus on specific influences such as gender and occupational stereotypes, work practices (such as working hours) and the occupational segregation women face in the DCI -Future research could focus on specific influences such as women’s responses to the tensions between gender and occupational identity

Further work to refine specific concepts

-Future research could refine the conceptualisation of several of the mechanisms, such as creativity -Future research could consider how the different properties of the ‘social identity’ dimension in the Sphere of Influence (SoI) model filter participation experiences -Future research could refine the element of temporality in the Sphere of Influence (SoI) model, and the emerging Acts of Agency theory

The Sphereof Influence (SoI) model and EventsofInteraction (EoI) model

-Further research could consider the relationships between the different contexts or settings of interaction, other than the workplace setting that this case study focused on (for example, future research may consider the relationships between the Community of Practice (CoP) and other settings) -Further data is required for a more detailed description of specific properties in the Sphere of Influence (SoI), such as policy and legislation

The Multi-Theory Scaffold(MTS)

-Future research could further develop the Multi-Theory Scaffold (MTS) as a methodological approach -Future research could further develop or modify the Human Agency Multi-Theory Scaffold (HAMTS) by considering other agency theorists, such as Emibayer and Mische (1998)

The FiveActs of Agency (FAA) model

-Future research could collect further empirical evidence to support the conceptualisation of agent-driven mechanisms -Further research could consider the relationships between the mechanisms in the Five Acts of Agency model -Future research could investigate if the mechanisms presented in the Five Acts of Agency are pertinent to the participation of other social identities, including gender (males), ethnicity and age

-Future research could consider the mechanisms identified in the FiveActs of Agency from a lifespan perspective

The emerging Acts of Agency theory

-Further theoretical work could develop the emerging theory by considering previous research that explores the linkages between Human Agency theory and Critical Realism

7.5 FINAL COMMENT

The research presented in this dissertation has contributed to the scholarly work

that considers the phenomenon of women’s participation in the Digital Content Industry

(DCI). The research addresses the gap in the literature by presenting an exploratory case

study, in which multiple sources of empirical data are analysed using three frameworks:

an analytical, theoretical, and ontological framework. The findings from the analysis

reveal there is a range of influences on women’s participation in the DCI. These

influences can be understood by using the original concept of agent-driven mechanisms.

I believe that this research offers an original research contribution. The personal

goal was to embrace creativity and imagination, while adhering to quality requirements.

As Mir & Mir (2002) remind us, we need the sociological imagination that Mills (1959)

argued for. Overall, the research establishes a body of knowledge about the phenomenon

of women’s participation within the Australian DCI, and contributes a valuable point of

reference for future endeavours that seek to understand the phenomenon. Furthermore,

the research may act as a catalyst for further research in this important area, or the

resultant models could prove to be useful tools for those researching participation in

other industries or contexts.

Danermark et al. (2002) remind us that the world is changing. On a personal note,

I welcome change regarding women’s participation in the DCI. I hope that in a decade

there will be no need to revisit the research problem; that the phenomenon of women’s

under-representation is consigned to being merely a historical record. Indeed, in the

course of undertaking the research, I have had stakeholders suggest that the issue of

women’s under-representation will eventually resolve itself; even participants note that

slowly, bit by bit more females are starting to join the industry (G2). The question I am

then left with is ‘How long will it take?’

As I sit and watch my seven-year-old daughter playing Fruit Ninja with great

gusto on her iPad, I notice the sticker on the back of the device.105 It reads, ‘Distributed

by the Office for Women (Queensland State Government) in March 2012 for

International Women’s Day’. The accompanying slogan proclaims that ‘Anything is

Possible’. Yes, anything is possible–eventually. I wonder if my daughter will have cause

to ask me in a decade, “Mum, why are there so few women working in multimedia and

game production organisations?”

105 The iPad was unveiled in January 2010 and became available to the public three months later. The sales of the iPad2 reached 4.69 million in March 2011 and approximately 50 million in January 2012 globally. Fruit Ninja was one of the highest grossing app games on ITunes in 2010.

References ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 325 of 401

References

Aaker, D.A., Day G.S. & Kumar V. (1998). Marketing Research. 6th ed. John Wiley & Sons, New York.

Abhayaratna, J. & Lattimore, R. (2006). Workforce Participation Rates - How Does Australia Compare? Staff Working Paper. World Wide Web Internet and Web Information Systems, (December).

Ackroyd, S. & Fleetwood S. (2000). Realism in contemporary organisation and management studies. In Realistic Perspectives on Management and Organisations, S. Ackroyd & S. Fleetwood (Eds), Routledge, London, 3-25.

ACS Australian Computer Society Ageism taskforce (2010). Improving age diversity in the ICT workforce, report, July 2010.

Adam, A., Howcroft, D. & Richardson, H., (2002). Guest Editorial, Information Technology and People, 15(2), 94-7.

Adam, A., Howcroft, D. & Richardson, H. (2004). A decade of neglect: reflecting on gender and IS, New Technology Work and Employment, 19(3).

Adams, D. (2006). E3 Booth Babes Banned. Companies may have to rely on actual games to grab our attention. IGN website. Online article http://au.ps2.ign.com/articles/683/683165p1.html

Adya, M. & Kaiser, K. M. (2005). Early determinants of women in the IT workforce: a model of girls’ career choices, Information Technology & People, Vol. 18(3), 230-259.

Agosto, D., Gasson, S., & Atwood, M. (2008). Changing Mental Models of the IT Professions: A Theoretical Framework. Journal of Information Technology Education, 7, 205-221.

Ahuja, M. K. (2002). Women in the Information Technology Profession: A Literature Review, Synthesis and Research Guide, European Journal of Information Systems, 11, 20-34. Retrieved August 10, 2005 from http://www.ingentaconnect.com/content/pal/0960085x/2002/00000011/00000001/3000417

AIMIA (2011). Working in Australia’s Digital Games Industry. Consolidation report. Australian Research Council Centre of Excellence for Creative Industries and Innovation (CCI) and Queensland University of Technology in partnership with the Games Developers Association of Australia.

AIMIA (Australian Interactive Media Industry Association). (2005). Digital Content Industry Roadmapping Study. Available at: http://www.aimia.com.au/i-cms_file?page=1455/Digital_Content_Roadmapping_Study_FINAL_AIMIA_Version.pdf

Alexandrowicz, C. L. (2005). Gendered Work and Women's Domestic Investment: A Second Look at the “Stalled Revolution”. Annual meeting of the American Sociological Association, Marriott Hotel, Loews Philadelphia Hotel, Philadelphia, PA Online Accessed 2009-05-25 from http://www.allacademic.com/meta/p22844_index.html

Allaire, Y., & Firsirotu, M. E. (1984). Theories of Organizational Culture. Organizational Studies, 5(3), 193-226

Allen, M. (2009). Historicising the Internet. Oxford Internet Institute Summer Doctoral Program 2009. http://netcrit.net/content/oiihistorisinginternet2009.pdf

Alvesson, M. (1996). Communications, Power and Organisation, Walter de Gruyter: Berlin.

Alvesson, M., & Karreman, D. (2007). Constructing mystery: Empirical matters in theory development.Academy of Management Review, 32, 1265–128.

American Association of University Women (AAUW). (2004). Under the Microscope: A Decade of Gender Equity Projects in the Science. Washington DC: AAUW Educational Foundation. www.aauw.org

American Game Career Guide. (2007). Game Developer magazine. Retrieved 30th August 2010 from website http://www.gamecareerguide.com/features/416/the_game_industry_salary_survey_2007.php?page=2

Andersen, P. H., & Kragh, H., (2009). Sense and sensibility: Two approaches for using existing theory in theory-building qualitative research, Industrial Marketing Management, 39, 49-55.

Andersen, V. & Gamdrup, P. (1990). Om problemformulering og projektarbejde. In Andersen, H. Videnskabsteori og metodelære. Frederiksberg, DK: Samfundslitteratur.

Anderson, N., Lankshear, C. & Klein, M. (2006). Women in Information and Communication Technology (ICT) careers. James Cook University, Education Queensland and Technology One. Retrieved September 19, 2006 from http://www.soe.jcu.edu.au/cgi-bin/rws3.pl?FORM=Industry_Survey_Women_Wizard

Archer, M. (1988). Culture and Agency: The Place of Culture in Social Theory, Cambridge: Cambridge University Press.

Archer, M. (1995). Realist Social Theory: the morphogentic approach. Cambridge: Cambridge University Press.

Archer, M. (1998). Introduction: realism in the social sciences, in Archer, M.S., Bhaskar, R., Collier, A., Lawson, T. & Norrie, A. (Eds), Critical Realism: Essential Readings, Routledge, London, 189-205.

Archer, M. (2000). Being Human: The Problems of Agency, Cambridge: Cambridge University Press.

Archer, M. (2003). Structure, Agency and the Internal Conversation. Cambridge: Cambridge University Press.

Archer, M. (2004). Emotions As Commentaries On Human Concerns, in J.H. Turner (ed.). Theory and Research on Human Emotions (Advances in Group Processes, Volume 21), Emerald Group Publishing Limited, pp.327-356

Archer, M. (2007). Making our Way Through the World, Cambridge University Press, Cambridge.

Archer, M. (2010). [1982] ‘Morphogenesis Versus Structuration: On Combining Structure and Action’, The British Journal of Sociology – The BJS: Shaping Sociology Over 60Years: 225–52. [Originally published in 1982 British Journal of Sociology 33(4): 455–83.]

Archer, M. S, & Bhaskar, R., Eds. (1998). Critical Realism: essential readings. Critical Realism--interventions. London, Routledge.

Australian Bureau of Statistics (ABS). (2005). Labour Force, Australia, (Cat. No. 6202.0). ABS, Canberra.

Australian Bureau of Statistics (ABS). (2007). Work in selected culture and leisure activities report, (Cat. No. 628 1.0). ABS, Canberra. http://www.abs.gov.au/ausstats/abs@.nsf/web%20pages/Citing%20ABS%20Sources#Publications

Australian Bureau of Statistics (ABS). (2008a). (Cat. No. 8515.0). Digital Game Development Services, Australia, ABS, Canberra. viewed 15 March 2009 http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/8515.0Main+Features12006-07?OpenDocument

Australian Bureau of Statistics (ABS). (2008b). Australian Culture and Leisure Classifications. (Cat No. 4902.0). ABS, Canberra. viewed 10 October 2008 http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/4902.02008%20(Second%20Edition).?OpenDocument

Australian Government. (2004). From reel to unreal: inquiry into the future opportunities for Australia’s film, animation, special effects and electronic games industries. House of Representatives Standing Committee on Communications, Information Technology and the Arts. Tabled: 21 June 2004.

Australian Government. (2005a). Inquiry into increasing participation in the workforce. Appendix E. Standing Committee on Employment, Workplace Relations and Workforce Participation Committee activities. http://www.aph.gov.au/house/committee/ewrwp/paidwork/report.htm

Australian Government. (2005b). Inquiry into increasing participation in the workforce. Appendix D. Standing Committee on Employment, Workplace Relations and Workforce Participation Committee activities. http://www.aph.gov.au/house/committee/ewrwp/paidwork/report.htm

Bacharach, S. B. (1989). Organizational theories: Some criteria for evaluation. Academy of Management Review, 14: 496-515.

Bakhshi, Hasan, Freeman, Alan, & Higgs, Peter L. (2012). A dynamic mapping of the UK's creative industries. NESTA, London.

Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1989a). Social Cognitive Theory. Annals of Child Development. 6, 1-60. (Vasta. R, ed). Greenwich, CT: Jai Press LTD.

Bandura, A. (1989b). Human Agency in Social Cognitive Theory. American Psychologist, 44, 1175-1184.

Bandura, A. (1991). Social cognitive theory of moral thought and action. In Handbook of Moral Behavior and Development. Kurtines WM & Gerwitz JL (eds). (Vol 1, p 45-103). Hillsdale, NJ: Erlbaum.

Bandura, A. (1997). Self-efficacy: The Exercise of Control. New York: W. H. Freeman.

Bandura, A. (1999a). A social cognitive theory of personality. In Pervin, L. & John, O. (Ed.), Handbook of Personality (2nd ed., 154-196). New York: Guilford Publications. (Reprinted in Cervone, D. & Shoda, Y. [Eds.] The coherence of personality. New York:Guilford Press.

Bandura, A. (1999b). Social Cognitive Theory: An agentic perspective. Asian Journal of Social Psychology, 2, 21-41.

Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1-26.

Bandura, A. (2002a). Growing Primacy of Human Agency in Adaptation and Change in the Electronic Era. European Psychologist, 7(1), 2-16.

Bandura, A. (2002b). Social Cognitive Theory in Cultural Context. Applied Psychology, 51(2), 269-290.

Bandura, A. (2002c). Social cognitive theory of mass communications. In J. Bryant, & D. Zillman (Eds.). Media effects: Advances in theory and research (2nd ed., 121-153). Hillsdale, NJ: Erlbaum. Bandura, A. (2003). On the psychosocial impact and mechanisms of spiritual modeling. International Journal for the Psychology of Religion, 13, 167-173.

Bandura, A. (2006). Toward a Psychology of Human Agency. Perspectives on Psychological Science, 1(2), 164-180.

Bandura, A. (2004). Selective exercise of moral agency. In Thorkildsen, T. A. & Walberg, H. J. (Eds.). Nurturing morality (37-57). Boston: Kluwer Academic.

Bandura, A., Barbaranelli, C., Caprara, G. V. & Pastorelli, C. (2001). Self-efficacy beliefs as shapers of children's aspirations and career trajectories. Child Development, 72, 187-206.

Barker, L. J., & Aspray, W. (2006). The state of research on girls and IT. In J. M. Cohoon, & W. Aspray (Eds.), Women and information Technology: Research on underrepresentation (pp. 3–54). Cambridge, MA: MIT Press.

Bazeley, P (2007). Qualitative data analysis with NVivo. Sage Publications, London.

Bazeley, P. (2009). Analysing Qualitative Data: More Than ‘Identifying Themes’ Submitted copy: Malaysian Journal of Qualitative Research, 2. Retrieved Oct 2010. http://www.researchsupport.com.au/More_than_themes.pdf

Beasley, B., & Standley, T. (2002). Shirts vs. skins: Clothing as an indicator of gender role stereotyping in video games. Mass Communication & Society, 5, 279-293.

Becker, G. S. (1985). Human capital, effort, and the sexual division of labor. Journal of Labor Economics, 3, Part 2, 33-58.

Beekhuyzen, J. & Clayton, K. (2004). ICT Career Perceptions: Not So Geeky? Proceedings of the 1st International Conference on Research on Women in IT, Kuala Lumpur, 28-30.

Beekhuyzen, J.P., Nielsen, S.H. & von Hellens, L. (2003). Challenging the dualisms in female perceptions of IT work. The Australian Journal of Information Systems, 10 (2), 105-114.

Berger, A. (1995) Cultural Criticism. London: Routledge.

Berger, P.L. & Luckmann, T. (1966), The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Doubleday, New York, NY.

Bertaux, D. (Ed.). (1981). Biography and society: The life history approach in the social sciences. Beverly Hills, CA: Sage.

Betz, N. E. & Hackett, G. (1997). Applications of self-efficacy theory to the career assessment of women. Journal of Career Assessment, 5, 383–402.

Bhaskar, R. (1978). A Realist Theory of Science. (2nd ed.). Brighton: Harvester Press.

Bhaskar, R. (1979). The Possibility of Naturalism, (1st Edition). Sussex: Harvester.

Bhaskar, R. (1986). Scientific Realism and Human Emancipation. London: Verso.

Bhaskar, R. (1989). Reclaiming Reality. London: Verso.

Bhaskar, R. (1991). Philosophy and the Idea of Freedom. Oxford: Blackwell.

Bhaskar, R. (1994). Dialectic: The Pulse of Freedom. London: Verso.

Bhaskar, R. (1998). The Possibility of Naturalism. (3rd ed.). London: Routledge.

Blake, K. (2011). A Woman in Games: A Personal Perspective, 1993 – 2010. International Journal.

Blum, L. & Frieze, C. (2005). The evolving culture of computing: Similarity is the difference. Frontiers: AJournal of Women Studies, 26, 110-125

Bowles, H., Babcock, L. & McGinn, K. (2005). Constraints and Triggers: Situational Mechanics of Gender in Negotiation, Working Paper Series rwp05-051, Harvard University, John F. Kennedy School of Government.

Brand, J., Borchard, J. & Holmes, K. (2009). IA9 Interactive Australia 2009. The Centre for New Media Research Bond University: Australia Retrieved August 2009 Available at: http://works.bepress.com/jeff_brand/12 Downloaded http://epublications.bond.edu.au/cgi/viewcontent.cgi?article=1098&context=hss_pubs

Brewer, M.B. & Gardner, W. (1996). Who is this,We?: Levels of collective identity & self representations. Journal of Personality & Social Psychology, 71, 83-93.

Brown, A., Grollmann, P., Tutschner, R., & PARTICIPA Project Consortium (2004). Participation in Continuing Vocational Education and Training: Results from the Regional and Sectoral Surveys. ITB-Arbeitspapiere Nr. 51. Bremen: ITB, Universität Bremen.

Brunner, C., Bennett, D., & Honey, M. (1998). Girl games and technological desire. In Cassell, J. & Jenkins, H. (1998). From Barbie to Mortal Kombat. Cambridge, MA: The MIT Press.

Bruns, A. (2008). Blogs, Wikipedia, Second Life, and beyond: from production to produsage. DigitalFormations, 45. Peter Lang, New York.

Bryant, C & Jary, D. (1991). Giddens' theory of structuration: a critical appreciation. Bryant, C. & Jary, D. (Editors). Routledge, London; New York.

Bryant, C. & Jary, D. (1997). Anthony Giddens: critical assessments, Routledge, (4 volumes).

Buchanan, D. A. & Bryman, A. (2009). The SAGE Handbook of Organizational Research Methods. SAGE Publications Ltd.

Bulkeley, D. (1995). Where to Recruit Your Next Rising Star, Datamation, 56-58.

Bury, R. (2011). She’s Geeky: The Performance of Identity among Women Working in IT. InternationalJournal of Gender, Science and Technology, Vol.3, No.1.

Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychology Review, 106(4), 676-713.

Calás, M.B. & Smircich, L. (2009). Feminist perspectives on Gender in Organisational Research: what is and is However to be. In Buchanan, D. & Bryman, A,(Eds.), SAGE handbook of organisational research methods. (246-269). SAGE, Great Britain: London.

Calhoun, C., Gerteis, J., Moody, J., Pfaff, S. and Virk, I. (eds). (2002). Contemporary Sociological Theory. Oxford: Blackwell.

Carlsson, S.A. (2006). Towards an Information Systems Design Research Framework: A Critical Realist Perspective, Proceedings of the First International Conference on Design Science in Information Systems and Technology (DESRIST 2006), 192-212.

Cassell, J. & Jenkins, H. (Eds). (1998). From Barbie to Mortal Kombat: Gender and Computer Games, Cambridge, MA: The MIT Press.

Castano, C. & Webster, J. (2011). Special Issue on Women in ICT: international research from a lifecourse perspective. International Journal of Gender, Science and Technology, 3(2).

Castells, M. (1998). The Information Age: Economy, Society and Culture. The Rise of Network Society, (1). Oxford: Blackwell.

Castells, M. (2001). The Internet Galaxy: Reflections on the Internet, Business and Society. Oxford: Oxford University Press.

Cater-Steel, A., & Cater, E. (2010). Women in Engineering, Science and Technology: Education and Career Challenges (pp. 1-384). doi:10.4018/978-1-61520-657-5

Cejka, M.A. & Eagly, A.H. (1999). Gender - stereotypic images of occupations correspond to the sex segregation of employment. Personality and Social Psychology Bulletin, 25, 413-42.

Chase, S. E. (2008). Narrative Inquiry: Multiple Perspectives, Approaches, Voices, 57-94. In Denzin, N.K. & Lincoln, Y. S. (Eds). Collecting and Interpreting Qualitative Materials (3rd ed.). Thousand Oaks, CA: Sage.

Christopherson, S. & van Jaarsveld, D. (2005). New Media after the Dot.com Bust. The Persistent Influence of Political Institutions on Work in Cultural Industries, International Journal of Cultural Policy 11(1).: 77-93.

Clegg, S.R. & Hardy, C. (1996). In Clegg, S.R. & Hardy, C. & Nord, W.R. (eds.). Handbook of Organization Studies. Sage, London, pp. 676-708.

Clayton, K. L., von Hellens, L. A., & Nielsen, S. H. (2009). Gender stereotypes prevail in ICT: A research review, SIGMIS-CPR '09, May 28-30, Limerick, Ireland, 153-158.

Cludts, S. (1999). Organisation Theory and the Ethics of Participation. Journal of Business Ethics, 23. 157-171. Retrieved from: http://www.jstor.org/stable/25074164

Cochran, L. (1997). Career counseling: A narrative approach. Thousand Oaks, CA: Sage Publication

McGrath- Cohoon, J. M. & Aspray, W. Editors. (2006). Women and Information Technology: Research on Under-Representation. MIT Press.

Collet, I.(2006). L’informatique a-t-elle un sexe?, Paris, L’Harmattan, .European Commission Best practices for even gender distribution in the 25 Member States in the domain of ICT, Brussels, DG Information Society and Media.

Collier, A. (1994). Critical Realism: An Introduction to Roy Bhaskar's Philosophy. London: Verso.

Connell, R.W. (1987). Gender and Power: Society, the Person and Sexual Politics. Sydney: Allen and Unwin.

Connell, R.W. (2002). Gender. Cambridge, Polity Press; Malden, Blackwell Publishers.

Connell, R.W., & Dowsett, G.W. (1993). Rethinking sex: Social theory and sexuality research: Temple University Press.

Connelly, J. (2001). Critical Realism and health promotion: effective practice needs an effective theory. Health Education. Research. 16, 115-120.

Conrad, C.F. & Serlin, R.C. (Eds.). (2006). The Sage Handbook for Research in Education: Engaging Ideas and Enriching Inquiry. 303-314. Thousand Oaks, California: Sage.

Consalvo, M. (2008). Crunched by passion: Women game developers and workplace challenges. In Y. B. Kafai, Y.B., Heeter, C., Denner, J. & Sun, J.Y. (Eds.), Beyond Barbie and Mortal Kombat. 177-192. Cambridge, MA: MIT Press.

Cookson, P.S. (1986). A framework for theory and research on adult education participation, AdultEducation Quarterly, 36(3), 130-141.

Corden, A. & Sainsbury, R. (2006). Using Verbatim Quotations in Reporting Qualitative Social Research: Researchers’ views, Social Policy Research Unit, University of York, York.

Courtney, L., Lankshear, C., Anderson, N. & Timm, C. (2009). Insider Perspectives vs. Public Perceptions of ICT: toward policy for enhancing female student participation in academic pathways to professional careers in ICT. Policy Futures in Education 7(1). 2009.

Craib, I. (1992). Modern social theory: from Parsons to Habermas. Hemel Hempstead: Harvester Wheatsheaf.

Craig, A. (2004). Redressing the Gender Imbalance in Australian Computing - Position Papers for the Oxford Internet Institute (OII). Policy Forum Retrieved June 1 2005 from http://www.oii.ox.ac.uk/resources/publications/OIIPP _20040617-WomenInIT_200407.pdf

Craig, A, Fisher, J., Forgasz, H. & Lang, C. (2011). Evaluation framework underpinning the digital divas programme, in unknown (ed.), ITiCSE 2011 : Proceedings of the 16th Annual Conference on Innovation and Technology in Computer Science, 313-317, Association for Computing Machinery, New York, N.Y.

Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage.

Creswell, J.W. (2003). Research Design, Qualitative, Quantitative, and Mixed Methods Approaches (2nd ed.). Thousand Oaks, CA: Sage.

Creswell, J.W. (2007). Qualitative Inquiry and Research Design: Choosing Among Five Traditions (2nd ed.). Thousand Oaks, CA: Sage.

Cunningham, S., Cutler, T., Ryan, M., Hearn, G. & Keane, M. (2003). Research and Innovation Systems in the Production of Digital Content and Applications. Content and Applications, Creative Industries Cluster Study Volume III. Commonwealth of Australia (DCITA). Canberra.

Cutler & Company Pty Ltd. (1996). From marketplace to marketspace, Australian Communications, February.

Dallimore, E. J. (2000). A Feminist Response to Issues of Validity in Research, Women's Studies in Communications, 23(2), 157-81.

Danermark, B., Ekstrom, M., Jakobsen, L., Karlsson, J. (2002). Explaining Society – Critical Realism in the Social Sciences. London: Routledge

Darke, P., Shanks, G. & Broadbent, M. (1998). Successfully completing case study research: combining rigour, relevance and pragmatism, Information Systems Journal, 8(4), 273-290.

DCITA. (2005). Australian Digital Content Industry Futures Plan, Prepared for DCITA (Department of Communications, Information Technology and the Arts) by the Centre for International Economics, March 2005. September 19, 2006 from http://www.dcita.gov.au/__data/assets/pdf_file/37474/Appendix_C.2_Australian_digital_content_futures.pdf

DCITA. (2006). Unlocking the potential. Digital Content Industry Action Agenda. Strategic Industry Leaders Group report to the Australian Government. Department of Communications, Information Technology and the Arts & Department of Industry, Tourism and Resources. Commonwealth of Australia, Canberra : Retrieved August 2007http://www.dbcde.gov.au/__data/assets/pdf_file/0006/37356/06030055_REPORT.pdf

Denzin, N. K., (1989a). The research act (3rd ed.). Englewood Cliffs, NJ: Prentice Hall.

Denzin, N. K., (1989b). Interpretive interactionism. Newbury Park, CA, Sage.

Denzin, N. K., (1989c). Interpretive biography. Newbury Park, CA: Sage

Denzin, N. K., & Lincoln, Y. S (Eds.). (2003). Landscape of Qualitative Research: Theories and Issues, (2nd ed). London: Sage.

Denzin, N. K., & Lincoln, Y. S (Eds.). (2008). Strategies of qualitative inquiry. Thousand Oaks, CA: Sage Publications.

Dervin, B., (1997). Given a context by any other name: Methodological tools for taming the unruly beast. In Vakkari, P., Savolainen, R. & Dervin, B. (Eds.), Information seeking in context 13-38. London: Taylor Graham.

Deuze, M., Martin, C., & Allen, C. (2007). The professional identity of gameworkers. Convergence, 13(4), 335–353.

DiMaggio, P. (1991). The Micro-Macro Dilemma in Organizational Research: Implications of Role-System Theory. In Huber, J. (Ed.). Macro-Micro Linkages in Sociology. 76-98. Newbury Park, CA: Sage.

Dobson, P.J., (1999). Approaches to Theory Use In Interpretive Case Studies – a Critical Realist Perspective. Proc. 10th Australasian Conference on Information Systems. Retrieved September 2nd, 2006, from http://www.vuw.ac.nz/acis99/Papers/PaperDobson-075.pdf

Dobson, P.J. (2001a). Longitudinal Case Research: A Critical Realist Perspective. Systemic Practice and Action Research, 14(3), 2001 Plenum Publishing Corporation.

Dobson, P.J. (2001b). The philosophy of Critical Realism - an opportunity for information systems research. Information Systems Frontier, 3(2), 199-201.

Dobson, P.J. (2002). Critical Realism and IS research - Why bother with philosophy? Information Research. Information research, 7(2). Retrieved Sept 2007 http://InformationR.net/ir/7-2/paper124.html

Dobson, P.J., Myles, J & Jackson, P., (2007). Making the Case for Critical Realism: Examining the Implementation of Automated Performance Management Systems, Information Resources Management Journal, 20(2). 140-152.

Donovan, J. (2001). Work Motivation. In Anderson, N. & Sinangil, H.K. (Eds.), Handbook of Industrial, Work and Organizational Psychology, 53-76, Sage.

Doolin, B. (1998). Information technology as disciplinary technology: Being critical in interpretive research on information systems [Special Issue on Interpretive Research in Information System]. Journal of Information Technology, 13(4), 301-312.

DTI (Department of Trade and Industry). (2005). Women in the IT Industry: Towards a Business Case for Diversity-Interim report. United Kingdom. Retrieved August 2nd, 2006, from http://www.dti.gov.uk/files/file9334.pdf

Duberley, J., Mallon, M., & Cohen, L. (2006). Exploring career transitions: accounting for structure and agency. Personnel Review, 35(3), 281-296. Retrieved from http://www.emeraldinsight.com/10.1108/00483480610656694

Dubois, A. & Gadde, L-E. (2002). Systematic combining – an abductive approach to case research, Journal of Business Research, 55(7), 553–560.

Easton G. (2000). Case research as a methodology for industrial networks: A Realist apologia, in Realist Perspectives on Management and Organisations in Fleetwood, S. & Ackroyd, S. (eds). Routledge, London and New York.

Easton, G. (2001). Case research as a method for industrial networks. A Realist apologia, in Realist Perspectives on Management and Organisations in Fleetwood, S. & Ackroyd, S. (eds). Routledge, London and New York. Viewed online 2009 http://books.google.com.au/

Easton, G. (2003). One Case Study is Enough. Academy of Marketing Conference 2003, Aston Business School.

Easton, G. (2005). Case study research: a Critical Realist approach, Lancaster University Management School.

Easton, G. (2009). Critical Realism in case study research, Industrial Marketing Management Retrieved: doi:10.1016/j.indmarman.2008.06.004

Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532-550.

Elliott, A.M. (2003). Critical visions: new directions in social theory, Lanham, USA: Rowman & Littlefield Publishers.

Emirbayer, M. & A. Mische. (1998). What Is Agency? American Journal of Sociology. 103(4), 962–1023.

Fairclough, N. (2005). Discourse Analysis in Organization Studies: The Case for Critical Realism, Organization Studies, 26(6).

Fayolle, A. (2011). Entrepreneurship and New Value Creation: The Dynamic of the Entrepreneurial Process. ISBN. 9781107402904.

Feyerband, P. (1999). For and Against Method: Including Lakatos's Lectures on Scientific Method and the Lakatos-Feyerabend Correspondence with Imre Lakatos, ISBN 0-226-46774-0, ISBN 0-226-46775-9

Flew, T. & Cunningham, S.D. (2010). Creative Industries After the First Decade of Debate. The Information Society, 26(2). Retrieved 30 March 2010 http://eprints.qut.edu.au/29852/2/29852.pdf

Flick, U. (2009). An Introduction to Qualitative Research (4th edition).London, England: SAGE.

Flores, F. & Gray, J. (2000). Entrepreneurship and the Wired Life: Work in the Wake of Careers, London: Demos

Florida, R. (2002). The Rise of the Creative Class. BasicBooks, New York.

Flyvbjerg, B. (2006). Five Misunderstandings about Case-Study Research. Qualitative Inquiry, 12(2), 219-245.

Fontana, A. & Frey, J. H. (1994). Interviewing: The Art of Science. In Denzin, N.K. & Lincoln, Y. S. (Eds). Handbook of Qualitative Research. (2nd ed.). 361-376. Thousand Oaks, CA: Sage.

Fox, S. (2009). Applying critical realism to information and communication technologies: A case study. Construction Management and Economics, 27(5), 465.

Fullerton, T., Fron, J., Pearce, C. & Morie, J. (2008). Getting girls into the game: Towards a “Virtuous Cycle.” In Kafai, Y. Heeter, C. Denner, J. & Sun, J. (Eds.), Beyond Barbie and Mortal Kombat: New perspectives on gender and gaming. Cambridge, MA: MIT Press.

Galliers, R. (1992). Choosing Information Systems Research Approaches. In Galliers, R. (Ed.), Information Systems Research - Issues, Methods and Practical Guidelines, 144-162. Alfred Waller Ltd.

Gallivan, M. J. (2004). Examining IT Professionals’ Adaptation to Technological Change: The Influence of Gender and Personal Attributes. The DATA BASE for Advances in Information Systems. Summer 2004. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

Gauntlett, D. (2001). The Worrying Influence of ‘Media Effects’ Studies, in Barker, M. & Petley, J. (eds), Ill Effects: The Media/Violence Debate, Routledge.

Geertz, C. (1973). Thick description: Toward an interpretive theory of culture, in The Interpretation of Cultures. New York: Basic Books, 3-30.

Geneve, A., (2011). Five Acts of Agency by women in the Australian Digital Content Industry. Console-ing Passions conference, 27-29 July, Adelaide, Australia 2011

Geneve, A., and Ganito, C. (2011). Women and Technology: Five Acts of Acts of Agency. In Araya, D. Breindl, Y, Houghton, T (Ed.) Nexus: New Intersections in Internet Research. Peter Lang:USA ISBN 978-1-4331-0970-6 pb.

Gerring, J. (2007). Is there a (viable). crucial-case method? Comparative Political Studies 40(3), 231-253.

Gibbs, G.R., Friese, S. & Mangabeira, W.C. (2002). The Use of New Technology in Qualitative Research. Introduction to Issue 3(2). of FQS [35 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(2), Art. 8, http://nbn-resolving.de/urn:nbn:de:0114-fqs020287.Gibbs, G & Taylor, C. (2005). Coding with Atlas.ti v.5. Retrieved September 13, 2005 from http://onlineqda.hud.ac.uk/Step_by_step_software/Atlas.ti/Atlas.ti_coding.php

Giddens, A. (1976). New Rules of Sociological Method, Basic Books, Hutchinson/New York.

Giddens, A. (1979). Central Problems in Social Theory: Action, Structure, and Contradiction in Social Analysis. Berkeley: University of California Press.

Giddens, A. (1984). The Constitution of Society. Polity Press, Cambridge.

Giddens, A. (1988). Social Theory Today,in Giddens, A. & Turner, J. (Editors), Social Theory Today, Stanford University Press, Stanford.

Giddens, A. (1991). Modernity and self-identity, Polity, Cambridge.

Giddens, A. (1992). The Transformation of Intimacy: Sexuality, Love and Eroticism in Modern Societies, Cambridge: Polity Press

Giddens, A. (1993). New Rules of Sociological Method: A Positive Critique of Interpretive Sociologies. New York: Basic Books.

Giddens, A. (1997). Sociology. 3rd Edition. Cambridge: Polity Press

Giddens, A., Duneier, M., Appelbaum, R.P. & Carr, D. (2008). Essentials of Sociology. 2nd Edition. New York: W.W. Norton & Company.

Giddens, A. & Pierson, C. (1998). Conversations with Anthony Giddens – Making Sense of Modernity. Polity Press, Cambridge.

Giddens, A. & Turner, J. (1988). Social Theory Today, Stanford University Press, Stanford.

Gill, R. (2002). Cool, Creative and Egalitarian? Exploring Gender in Project-Based New Media Work in Europe. Information, Communication and Society, 5(1), 70-89.

Gill , R. (2006). Gender and the Media. Cambridge: Polity Press.

Gill, R. (2007). Technobohemians or the new Cybertariat? Understanding contemporary new media work. Institute of Network Cultures, Amsterdam.

Gill, R. & Pratt, A. (2008). In the Social Factory? Immaterial Labour, Precariousness and Cultural Work. Theory, Culture & Society, 25(7-8), 1-30.

Gill, R. (2010) "Life is a pitch": managing the self in new media work. In Mark Deuze (ed.) Managing MediaWork. London (etc) Sage.

Gioia, D. & Pitre, E. (1990). Multiparadigm Perspectives on Theory Building, Academy of Management Review 15. 584–602.

Glaser, B.G. (1992). Emergence vs. Forcing: Basics of Grounded Theory, Sociology Press, Mill Valley, CA.

Gorski, P.S. (2008). The ECPRES Model: A Critical Realist Approach to Causal Mechanisms in the Social Sciences. in Wittrock, B. & Hedström, P. (Editors). The Frontiers of Sociology, 147-194. Leiden: Brill.

Gourdin, A. (2005). Game developer demographics: an exploration of workforce diversity. InternationalGame Developers Association, [online], http://www.igda.org/diversity/IGDA_ DeveloperDemographics_Oct05.pdf, accessed 24 January 2008.

Green, J. & Thorogood, N. (2009). Qualitative methods for health research (2nd edition). Thousand Oaks, CA: Sage.

Gregg, M. (2008). Testing the Friendship: Feminism and the Limits of Online Social Networks, FeministMedia Studies, 8(2): 197–223

Gregg, M. and Seigworth, G.J.eds. (2010). The Affect Theory Reader. Duke University Press. 416 pp

Gregor, S. (2002). A theory of theories in information systems. In Gregor, S. & Hart, D. (Eds.). Information Systems Foundations: Building the Theoretical Base. Australian National University, Canberra, 1-20.

Gregor, S. (2006). The nature of theory in information systems. Management Information Systems Quarterly, 30(3), 611-642.

Greig, A., Lewins, F. & White, K. (2003). Inequality in Australia. Cambridge: Cambridge University Press.

Griffiths, M, Keogh, C., Moore, K., Richardson, H., & Tattersall, A. (2006). Inclusion through the ages? Gender, ICT workplaces and lifestage experiences. In Kendall, J., Howcroft, D., et al. (Editors). Social inclusion: societal and organisational implications for information systems. Springer, New York, 153-168.

Griffiths, M., & Moore, K. (2010). Disappearing Women: A Study of Women Who Left the UK ICT Sector? Technology Management, 5(1).

Griffiths, M., Keogh, C., Moore, K., Tattersall, A. & Richardson, H. (2005). Managing Diversity or Valuing Diversity: Gender and the IT Labour Market. In Neiderman, F. & Ferratt, T. (Editors). IT Workers: Human Capital Issues in a Knowledge-Based Environment, Hershey, PA, USA: Information Science Publishing.

Griffiths, M., Moore, K., & Richardson, H. (2007). Celebrating Heterogeneity: A survey of female ICT professionals in England. Information, Communication & Society, 10(3), 338-357.

Griffiths, M., Moore, K., Burns, B., & Richardson, H. (2007). The Disappearing Women: North West ICT Project Final Report. Funded by the European Social Fund (ESF). Spring. Retrieved Feb 2010 http://usir.salford.ac.uk/9312/1/Griffiths_DW.pdf

Grotz, B. (1999). Contexts of Collaboration. In Cognition, Agency and Rationality: Proceedings of the Fifth International Colloquium on Cognitive Science, Kepa Korta, Sosa, E. & Arrazola, X. (Eds.), Philosophical Studies Series 79, Dordrecht/Boston: Kluwer Academic Press.

Gruys, M., Munshi, N., & Dewett, T. (2011). When antecedents diverge: Exploring novelty and value as dimensions of creativity, Thinking Skills and Creativity, In Press, Corrected Proof available online 5 February 2011, ISSN 1871-1871, DOI: 10.1016/j.tsc.2011.01.005. (http://www.sciencedirect.com/science/article/pii/S1871187111000174).

Guba, E. G. & Lincoln, Y. S. (1989). Fourth generation evaluation. Newbury Park, CA: Sage.

Guba, E. G. & Lincoln, Y. S. (2005). Paradigmatic controversies, contradictions and emerging confluences. In Denzin, N. & Lincoln, Y. S. (Eds.). The Sage Handbook of Qualitative Research, 8, 191-215, Sage Publications.

Gurer, D. & Camp, T. (2002). Investigating the incredible shrinking pipeline for women in computer science (Final report). National Science Foundation Project 9812016. Retrieved on 6 May 2006 from www.acm.org/women/ pipeline-finalreport_ver_2.doc

Haukka, S (2009). From education to work in Australia’s creative digital industries: the opinions and practices of aspiring creatives in the creative industries, 60Sox report vol. 1, QUT, Brisbane, viewed 19 May 2013, <http://eprints.qut.edu.au/29795/1/60Sox_Survey_Report_Volume_June_2009.pdf> Accessed 19 June 2011

Haukka. S, and Brow,B. (2010). From education to work in Australia’s Creative Digital Industries: Comparing the opinions and practices of employers and aspiring creatives (60Sox Report Volume 2, January 2010), available at http://www.apo.org. au/research/education-work-australias-creative-digital-industries.. Accessed 19 June 2011.

Haines, L. (2004). Why are there so few women in games? Research for Media Training North West.

Halford, S., Savage M. & Witz, A. (1997). Gender, careers and organizations. Current developments in banking, nursing and local government, Macmillan.

Hammersley, M. (1995). Theory and evidence in qualitative research. Quality and Quantity, 29(1). 55-66.

Hammersley, M. (2009). Why critical realism fails to justify critical social research. MethodologicalInnovations Online, 4(2). 1–11.

Hammersley, M. & Atkinson, P. (1983). Ethnography: Principles in practice, Tavistock: London.

Hammersley, M. & Atkinson, P. (1995). Ethnography: Principles in practice, 2nd Ed. Routledge, London and New York.

Hammersley, M. & Gomm, R. (2000). Introduction. In Gomm, R. Hammersley, M. & Foster, P. (Editors). Case study method. London: Sage. 116.

Harding, S. (2004). The Feminist Standpoint Theory Reader: Intellectual and Political Controversies, Routledge, New York, NY,

Harrison, D. & Easton, G. (2004). Temporally embedded case comparison in industrial marketing research. In Critical Realist Applications in Organisation and Management Studies, Fleetwood, S. Ackroyd, S. (Editors). Routledge: London, U.K. 194–210.

Hatch, M.J. (1997). Organization Theory: Modern Symbolic and Postmodern Perspectives. Oxford: Oxford University Press.

Hayes, E. (2008). Girls, gaming, and trajectories of technological expertise, in Kafai, Y.B., Heeter, C., Denner, J., & Sun, J. (eds.). Beyond Barbie and Mortal Kombat: New Perspectives on Gender, Games, and Computing, Boston, MIT Press.

Healy, K. (1998). Conceptualising Constraint: Mouzelis, Archer and the Concept of Social Structure. Sociology, 32(3), 509-522.

Healy, M. & Perry, C. (2000). Comprehensive criteria to judge validity and reliability of qualitative research within the realism paradigm. Qualitative Market Research, 3(3), 118-126. Retrieved December 6, 2009, from ABI/INFORM Global.

Heeter, C., Egidio, R., Mishra, P., Winn, P. & Winn, J. (2009). Alien Games: Do Girls Prefer Games Designed by Girls? Games and Culture, 4(1), 74-100.

Heidegger, M. (1962). Being and Time, Basil Blackwell: Oxford.

Higgs, P. & Cunningham, S. (2008), Creative Industries Mapping: Where have we come from and where are we going?, Creative Industries Journal, 1(1), 7–30.

Hitlin, S. & Elder, Jr. G.H. 2007. Agency: An Empirical Model of an Abstract Concept. In Advances in Life Course Research, 11, edited by Macmillan, R. Netherlands: Elsevier/JAI Press. 33–67

Hodgson, G. (2004). Some claims made for critical realism in economics: two case studies, Journal of Economic Methodology, Taylor & Francis Journals, 11(1), 53-73.

Hofstede, G. (1993). Cultural Constraints in Management Theories, Academy of Management Executive, February.

Hogg, M. A., & Terry, D. J. (Eds.). (2001) Social identity processes in organizational contexts. Philadelphia: Psychology Press.

Holland, D. & Lave, J. (Eds.) (2001). History in Persons: Enduring struggles, contentious practice, intimate identities. Sante Fe, NM:School of American Research, Oxford.

Holstein, J.A., & Gubrium, J. F. (1995). The Active Interview. Qualitative Research Methods, 37. London: Sage Publications.

Hoonakker, P.L.T., Carayon, P. & Schoepke, J. (2006). Discrimination and Hostility Toward Women and Minorities in the IT Workforce. In Encyclopaedia of Gender and Information Technology (E. Trauth, editor). Hersey, PA: Idea Group Inc, 207-215.

Houghton, J. (2001). Impact of the ICT industry in Australia. Report for Australian Computer Society. Retrieved Sept 17 2007 from http://www.acs.org.au/notices/itcimpact.pdf

Howcroft, D., & Trauth, E.M. (2005). Handbook of Critical Information Systems Research: Theory and Application, Edward Elgar, Cheltenham,

Howcroft, D., & Trauth, E.M. (2008). The Implications of a Critical Agenda in Gender and IS Research, Information Systems Journal Special Issue: Exploring the Critical Agenda in IS Research. 18(2): 185-202.

Huff, C. (2002). Gender, software design, and occupational equity. Women and Computing [Special issue]. ACM SIGCSE Bulletin, 34(2), 112 – 115.

Huntemann, N. (2007). Pink Slips for Booth Babes?: No Way! Re-train and Re-skill! Special Issue: Video Games, Featured Articles, Volume 5. Retrieved 30th Aug 2010 from http://flowtv.org/2007/02/pink-slips-for-booth-babes-no-way-re-train-and-re-skill/

Hyde, J. S. (1990). Meta-Analysis and the Psychology of Gender Differences. Signs, 16(1), 55-73.

IBSA (2011), Environment Scan. Information and Communications Technology Industries. Department of Education, Employment and Workplace Relations (DEEWR). Australian Government

IGDA (2005a). Game Developer Demographics: An Exploration of Workforce Diversity Comments on Diversity. Downloaded http://archives.igda.org/diversity/IGDA_DeveloperDemographics_Oct05.pdf

IGDA (2005b). Game Developer Demographics: An Exploration of Workforce Diversity supplemental resource of raw data. http://www.igda.org/sites/default/files/IGDA_Comments-on-Diversity_Jul05.pdf

Ilie, V., Van Slyke, C., Green, G. & Lou, H. (2005). Gender Differences in Perceptions and Use of Communication Technologies: A Diffusion of Innovation Approach, Information Resource Management Journal, 18(3), 16-31.

Information Technology Association of America (ITAA). (2005). Untapped talent: Diversity, competition, and America’s high tech future–executive summary. Retrieved from http://www.itaa.org/eweb/upload/execsummdr05.pdf

Ivory, J. (2006). Still a Man's Game: Gender Representation in Online Reviews of Video Games. MassCommuni Ilie cation & Society, 9(1), 103-114.

James, W. (1907). Pragmatism: A New Name for Some Old Ways of Thinking. New York, 52-61.

Jansz, J. (2005). The emotional appeal of violent video games for adolescent males. Communication Theory, 15(3), 219-241.

Jansz, J., & Martis, R. (2003). The representation of gender and ethnicity in digital interactive games. In Copier, M. & Raessens, J. (Eds.), Level Up: Digital Games Research Conference 260-269. Utrecht: Utrecht University.

Jenkins, H. & Cassell, J. (2008). From Quake Grrls to Desperate Housewives: A Decade of Gender and Computer Games. In Kafai, Y., Heeter, C., Denner, J. & Sun, J. (eds.). Beyond Barbie and Mortal Kombat: New Perspectives on Gender and Gaming. Cambridge, MA: MIT Press. Retrieved http://www.soc.northwestern.edu/justine/publications/Jenkins_Cassell%20BBMK_Forward.pdf

Jeppson, S. (2005). Critical Realism as a Approach to Unfolding Empirical Findings: Thoughts on Fieldwork in South Africa on SMEs and Environment. The Journal of Transdisciplinary Environmental Studies, 4(1).

Jeppson, A. & Perl, T. (2002). Priming the pipeline. Inroads SIGCSE Bulletin, 34(2), 36-39.

Joas, H. (1993). Pragmatism and Social Theory, Chicago, University of Chicago Press.

Joas, H. (1997). The Creativity of Action, Chicago, University of Chicago Press.

Johnson, R. B. & Onwuegbuzie, A.J. (2004). Mixed Methods Research: A Research Paradigm Whose Time Has Come. Educational Researcher, 33(7), 14-26.

Jones, M.R. (1999). Structuration theory. In Currie, W.J. &. Galliers, R.D. (Eds.). Re-thinking Management Information Systems, 103-135. Oxford: Oxford University Press.

Jones, M. & Karsten, H. (2003). ‘Review: Structuration theory and information systems research’. WP 11/03. Judge Institute Working Papers. Cambridge: University of Cambridge.

Joshi, K.D. & Kuhn, K. (2005). Examining the masculinity and femininity of critical attributes necessary to succeed in IT. Proceedings of the 2005 ACM SIGMIS CPR conference on Computer Personnel Research, April 14-16, 2005, Atlanta, Georgia, USA (pp. 32-35). New York, NY, USA : ACM Press. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

Joshi, K.D., Kvasny, L., McPherson, S., Trauth, E., Kulturel-Konak, S. & Mahar, J. (2010). Choosing IT as a career: exploring the role of self-efficacy and perceived importance of IT skills, ICIS 2010 Proceedings. Paper 154. http://aisel.aisnet.org/icis2010_submissions/154

Kanter, R. (1977) Men and women of the corporation. New York: Basic Books.

Kappos, A. & Rivard, S. (2008). A Three-Perspective Model of Culture, Information Systems, and Their Development and Use, MIS Quarterly, 32(3); 601-634.

Kelle, U. (1995) Capabilities for "Theory Building" and "Hypothesis Testing" in Software for Computer-aided Qualitative Data Analysis. Retrieved September 13, 2005 from http://caqdas.soc.surrey.ac.uk/kellecapabilities.pdf

Kerr, A. (2006). The Business and Culture of Digital Games, Game work and Game Play. Sage, London.

Kincheloe, J. L. (2001). Describing the Bricolage: Conceptualizing a New Rigor in Qualitative Research. Qualitative Inquiry, 7(6), 679-692.

Klawe, M., (2002) Girls, boys, and computers. SIGCSE Bulletin, 34(2), 16-17

Klein, H. K. & Myers, M.D. (1999). A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems, MIS Quarterly, Special Issue on Intensive Research. 23(1), 67-93.

Klein, K. J., Cannella, A. & Tosi, H. (1999). Multilevel theory: Challenges and contributions. Academy of Management Review, 24, 243-248.

Koesten, J., Miller, K.I. & Hummert, M.L. (2002). Family communication, self-efficacy and white adolescent females' risk behavior. Journal of Family Communication, 1; 7-27.

Krotoski, A. (2004). Chicks and joysticks: An exploration of women and gaming. White paper Entertainment and Leisure Software Publishers Association (ELSPA)

Kvasny, L. Joshi, K.D. & Trauth, E.M. (2011). The Influence of Self-Efficacy, Gender Stereotypes and the Importance of IT Skills on College Students’ Intentions to Pursue IT Careers, Proceedings of the 2011 iConference (Seattle, WA).

Kvasny L. & Trauth, E. (2002). The Digital Divide at Work and at Home: Discourses about Power and Underrepresented Groups in the Information Society. In Whitley, E. Wynn, E. & DeGross, J. (Eds.), Global and Organizational Discourse About Information Technology, NewYork: Kluwer Academic Publishers. 273-294.

Kvasny, L., Trauth, E.M. & Morgan, A. (2009). Power Relations in IT Education and Work: The Intersectionality of Gender, Race and Class, Journal of Information, Communication and Ethics in Society Special Issue on ICTs and Social Inclusion. 7(2/3), 96-118.

Lakatos, I. (1978). The Methodology of Scientific Research Programmes, ed. J. Worrall & G. Currie. Cambridge: Cambridge.

Lang, C. (2003). Gender and the geek factor: Why don't women do IT? Traffic, 3, 79-91.

Lang, C. (2007a). Twenty-first Century Australian Women and IT: Exercising the power of choice, Computer Science Education, 17(3), 215-226. Retrieved April 20, 2008, from http://www.informaworld.com.ezp02.library.qut.edu.au/10.1080/08993400701538120

Lang, C. (2007b). Factors that shape student decision-making related to information technology study and career choices: a gendered analysis. PhD Thesis.

Lang, C., Craig, A., Fisher, J., & Forgasz, H. (2010). Creating Digital Divas – Scaffolding Perception Change Through Secondary School and University Alliances. Computer, 38-42.

Lanzalotto, J. (2007). How to narrow IT's gender gap. Winning the hearts and minds of the industry's top female talent could be easy. In Computerworld. Retrieved 30th August 2010 from http://www.computerworld.com.au/article/61206/how_narrow_it_gender_gap/#closeme

Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation. Cambridge: Cambridge University Press.first published in 1990 as Institute for Research on Learning report 90-0013.

Lawson, T. (1997). Economics and Reality. London: Routledge.

Layder, D. (1993). New Strategies in Social Research. Polity Press, Cambridge, UK.

Layder, D. (1998). Sociological practice: linking theory and social research. Sage, London.

Layder, D. (2006). Understanding social theory, Sage.

Lee, A.S. & Baskerville, R. L. (2003). Generalizing Generalizability in Information Systems Research. Information Systems Research, 14(3), 221–243.

Leedy, P. D., & Ormrod, J. E. (2005). Practical research: Planning and design. Upper Saddle River, NJ: Prentice-Hall.

Lent, R.W., Brown, S.D., & Hackett, G. (1994). Monograph: Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behaviour, 45, 79-122.

Lent, R.W., Brown, S.D., & Hackett, G. (1996). Career development from a social cognitive perspective. In Brown, D. & L. Brooks (Eds.), Career choice and development. San Francisco, CA: Jossey-Bass Inc. 373-421.

Lent, R.W., Brown, S.D., Schmidt, J., Brenner, B., Lyons, H. & Treistman. (2003). Relation of contextual supports and barrier, Ds to choice behavior in engineering majors: Test of alternative social cognitive models. Journal of Counseling Psychology 50(4), 458–65.

Levy, Y. & Ellis, T.J. (2006). A Systems Approach to Conduct an Effective Literature Review in Support of Information Systems Research. Science Journal, 9.

Lewis, M.W. & Grimes, A.J. (1999). Metatriangulation: Building Theory from Multiple Paradigms, Academy of Management Review, 24(4), 672-690.

Lewis, P. & Simpson, R. (2007). Gendering Emotions in Organizations. Basingstoke: Palgrave.

Lincoln, Y.S., & Guba, (1985). Naturalistic inquiry. Newbury Park, CA: Sage.

Lincoln, Y.S., & Guba, (2000). Paradigmatic controversies, contradictions, and emerging confluences. In Denzin, N.K. & Lincoln, Y.S. (Eds.), Handbook of qualitative research (2nd edition), 163-188. ThousandOaks, CA: Sage.

Lippa, R. (2002). Gender, Nature, and Nurture. Mahwah, NJ: Lawrence Erlbaum Associates.

Lind, M., & Goldkuhl, G. (2006). How to develop a multi-grounded theory: The evolution of a business process theory. Australasian Journal of Information Systems, 13 (2), 1-10.

Llewelyn, S. (2003). What counts as "Theory" In qualitative management and accounting research? Introducing five levels of theorising. Accounting, Auditing & Accountability Journal, 16(4); 662-708.

Lloyd, M. (2009). Geek goddesses go to school: A study of female preservice teachers in information technology courses. In BERA 2009 Annual Conference, September 2-5, University of Manchester, Manchester.

Lofland, J. & Lofland, L. (1995). Analyzing social settings, 3rd Edition. Belmont, Cal: Wadsworth.

Lòpez, J. & Potter, G. (Eds) (2001). After Postmodernism: An Introduction to Critical Realism. Athlone, London.

Luck, L., Jackson, D. & Usher, K. (2006). Case study: a bridge across the paradigms. Nursing Inquiry, 13, 103–109.

Macknight, J. (2001). Nerdy image and blokish behaviour drive women away from IT. Computer Weekly, March 8.

Mackrell, D. & Nielsen, S. (2007). Structuration Theory in Information Systems Research: Relevance and Rigour from a Pluralist Research Approach. Proceedings of the 18th Australasian Conference on Information Systems, University of Southern Queensland, Queensland, December

Major, D.A., Davis, D.D., Sanchez-Hucles, J.V., Germano, L.M. & Mann, J. (2006). IT Workplace Climate for Opportunity and Inclusion. In Trauth, E.M. (Ed.), Encyclopedia of Gender and Information Technology, 856-862.

Margolis, J. & Fisher, A. (2002). Unlocking the Clubhouse. Cambridge, MA: MIT.

Martin, C. & Deuze, M. (2009). The Independent Production of Culture: A Digital Games Case Study. Games and Culture, 4(3), 276-295.

Maslow, A.H. (1962). Notes on Being-Psychology Journal. Humanistic Psychology, 2,47-71.

Mathews, M. (2008). Science and Innovation Policy and the New (and Old). Economics of Creativity, Vol, 1 (1). Cultural Science. ISSN 1836-0416. Retrieved 19th August 2009 online http://cultural-science.org/journal/index.php/culturalscience/article/view/9

May C.R, Mair F, Finch T, Macfarlane A, Dowrick C, Treweek S, Rapley T, Ballini L, Ong B.N, Rogers A, Murray E, Elwyn G, Légaré F, Gunn J, Montori V.M. (2009) Development of a theory of implementation and integration: normalization process theory. Implement Sci. (4) 29.

Maxwell, J. A. (1992). Understanding and validity in qualitative research. Harvard Educational Review, 62(3), 279-300.

Maxwell, J.A. (2005). Qualitative research design: An interactive approach (2nd ed). Thousand Oaks, CA: SAGE Publications.

McGrath, J., & Kelly, J. (1986). Time and human interaction: towards a social psychology of time, New York: Guildford Press.

McGrath-Coohan , J.M. & Aspray, W. (Eds). (2006). Women in Information Technology: Research on Under-representation. Cambridge, MA: MIT Press.

McLennan, G. (1989). Marxism, Pluralism and Beyond, Polity, 165-197.

Melymuka, K. (2000). Glass ceilings & clear solutions. Computerworld. 34(22). 53.

Merriam, S.B. (1998). Qualitative research and case study applications in education. San Francisco: Jossey-Bass.

Merton, R. (1967). On Theoretical Sociology. Free Press, New York.

Meyerson, D.E., & Fletcher, J.K. (2000). A modest manifesto for shattering the glass ceiling. HarvardBusiness Review, 78(1), 126-136.

Miles, M.B., & Huberman, A.M. (1994). Qualitative data analysis: an expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage Publications, Ltd.

Mills, A., Durepos.G. & Wiebe, E. (2009). Naturalistic generalization. Encyclopedia of Case Study Research. Edited by Mills, A.J., Durepos, G. & Wiebe, E. Thousand Oaks, CA: Sage Publications.

Mills, C.W. (1959). The Sociological Imagination, Oxford University Press, London.

Mingers, J. (1997). Combining research methods in information systems: multiparadigm methodology, Proceedings of the Fifth European Conference on Information Systems, Cork, Ireland, 19-21 June, 760–776.

Mingers, J. (2000). The contribution of Critical Realism as an underpinning philosophy for OR/MS and systems, Journal of the Operational Research Society, 51(11), 1256-1270.

Mingers, J. (2001). Combining IS research methods: Towards a pluralist methodology. Information Systems Research, 12(3), 240-259.

Mingers, J. (2002). Realizing information systems: Critical realism as an underpinning philosophy for information systems, in International Conference on Information Systems, (ed, DeGross, J). Association of Information Systems, Barcelona, 295-303.

Mingers, J. (2003). The paucity of multimethod research: review of the IS literature. Information Systems Journal, 13(3), 233-249.

Mingers, J. (2004). Realizing information systems: Critical Realism as an underpinning philosophy for information systems. Information and Organization, 14(2), 87-103.

Mingers, J. (2006). Realising Systems Thinking: Knowledge and Action in Management Science. New York: Springer.

Mir, R. & Mir, A. (2002). The Organizational Imagination: From Paradigm Wars to Praxis, Organizational Research Methods, 5, 105-125.

Montano, S. & Szmigin, I. (2004). Case study research: A Critical Realist approach. (Eds). Brown, A & Remenyi, D. Proceedings of the 4th European Conference on Research Methods: In Business and Management Studies. Universite Paris Duaphine, Paris.

Moore, A. K., Griffiths, M., & Richardson, H. (2005a). Moving in, Moving up, Moving out? A Survey of Women in ICT. Information Systems, 1-28.

Moore, A. K., Griffiths, M., & Richardson, H. (2005b). Gender on the Agenda: flexible working and work-life balance in the UK ICT industry, 6th International Women in Computing (WIC). Conference 14-15 July 2005, University of Greenwich London UK.

Morgan, A.J., Quesenberry, J.L. & Trauth, E.M. (2004). Exploring the Importance of Social Networks in the IT Workforce: Experiences with the ‘Boy’s Club.’ Proceedings of the Americas Conference on Information Systems (New York, August).

Morgan, G. & Smircich, L. (1980). The case for qualitative research. Academy of Management Review, 5(4), 491-500.

Morse, J.M. (1994). Designing funded qualitative research. In Denzin, N.K. & Lincoln, Y.S. (Eds.), Handbook of qualitative research, 220-233. Thousand Oaks, CA: Sage.

Morse, J. & Richards, L. (2002). Read me first for a user's guide to qualitative methods. London, Sage Publications.

Mouzelis, N. (1995). Sociological Theory: What Went Wrong? New York: Routledge.

Mutch, A. (1997). Critical Realism and information systems: an exploration. The 7th Annual BIT Conference, Manchester.

Mutch, A. (2007). Reflexivity and the Institutional Entrepreneur: A Historical Exploration. OrganizationStudies, 28, 1123-1140.

Myers, M.D. (1994). Dialectical Hermeneutics: a Theoretical Framework for the Implementation of Information Systems, Information Systems Journal, 5, 51-70.

Myers, M.D. (1997). Interpretive research in information systems. In Mingers, J. & Stowell, F.A. (Eds.), Information systems: An emerging discipline? London: McGraw-Hill. 239-266.

Myers, M.D. & Newman, M. (2007). The qualitative interview in IS research: Examining the craft, Information and Organization, 17(1), 2-26.

Negroponte, N. (1999). Being Digital. New York: Knopf.

Nielsen Online Consumer report (2010) The Australian Online Consumer Report. Retrieved March 2011 http://www.ad-tech.com/sydney_uploads/nielsen_WhitePaper2011.pdf

Neuman, W. L. (1997). Social research methods: Qualitative and quantitative approaches. Boston: Allyn and Bacon.

Neuman, W. (2000). Social Research Methods: Qualitative and quantitative approaches. Needham Heights: Allyn & Bacon.

Newmarch, E., Taylor-Steele, S. & Cumpston, A. (2000). Women in IT- What are the Barriers. Network of Women in Further Education Conference “Net Gains: Women, Information Technology and Emerging Issues”, March 22. Department of Education, Science and Training. Australia.

Olsen, W. (2008). Realist ontology and epistemology for rural research, September 2008 BWPI Working Paper 53. World, (September), 1-26.

Onwuegbuzie, A.J. & Leech, N.L. (2005a). The role of sampling in qualitative research. Academic Exchange Quarterly, 9(3), 280-284.

Onwuegbuzie, A. J., & Leech, N. L. (2005b). A Typology of Errors and Myths Perpetuated in Educational Research Textbooks, Current Issues in Education [On-line], 8(7). Available: http://cie.ed.asu.edu/volume8/number7/

Onwuegbuzie, A. J. & Leech, N. L. (2007). Sampling designs in qualitative research: Making the sampling process more public. The Qualitative Report, 12(2), 238-254.

Orlikowski, W.J. & Baroudi, J.J. (1991). Studying Information Technology in Organizations: Research Approaches and Assumptions, Information Systems Research (2), 1-28.

Orlikowski, W.J. & Robey, D. (1991). Information Technology and the Structuring of Organizations.Information Systems Research, 2 (2), 143-169.

Oswald, D. L. (2008). Gender stereotypes and women's reports of liking and ability in traditionally masculine and feminine occupations. Psychology of Women Quarterly, 32, 196-203. Retrieved May 22nd 2008, http://www.ingentaconnect.com/content/bpl/pwqu/2008/00000032/00000002/art00008"

Panteli, N., Stack, J., Atkinson, M. & Ramsay, H. (1999). The status of women in the UK IT industry: an empirical study, European Journal of Information System, 8, 170-82.

Panteli, N. (2005). Women, Work & IT: Reproducing gender inequality: Segregation and career paths in information technology jobs in Australia. Proceedings of the Women Work and IT conference: Contemporary perspectives on the reproduction of gender inequality in employment, June 23-24,

2005, Brisbane, Australia. School of Political Science and International Studies: University of Queensland.

Parberry, I., Kazemzadeh, M. & Roden, T. (2006).The art and science of game programming, Proceedings of the 37th SIGCSE technical symposium on Computer science education, March 03-05, 2006, Houston, Texas, USA.

Parker, J. (2000). Structuration, Buckingham, Open University Press.

Patton, M. (2002). Qualitative research and evaluation methods (3rd ed.). London:Sage.

Patton, M.Q. (1990). Qualitative Evaluation and Research Methods, Newbury Park, Sage Publications.

Patton, W., Bartrum, D.A. & Creed, P.A. (2004). Gender differences for optimism, self esteem, expectations and goals in predicting career planning and exploration in adolescents. InternationalJournal for Educational and Vocational Guidance, 4(2-3), 193-209.

Pawson, R., Greenhalgh, T., Harvey, G. & Walshe, K. (2004). Realist synthesis: an introduction. Manchester: University of Manchester, Faculty of Social Sciences. (ESRC Research Methods Programme Paper 2/2004). Retrieved 28 December, 2010 from http://www.ccsr.ac.uk/methods/publications/RMPmethods2.pdf (Archived by WebCite® at http://www.webcitation.org/5vJiVkJ8A).

Pearl, A., Pollack, M.E., Riskin, E., Thomas, B., Wolf, E. & Wu, A. (1990). Becoming a computer scientist: A report by the ACM Committee on the status of women in computing science. Communications of the ACM, 33(11), 47-57.

Pease, A. & Pease, B. (1998).Why men don’t listen and women can’t read maps: How we're different and what to do about it. New York: Broadway Books

Peirce, C.S. (1955). Abduction and induction. In Justus Buchler (Ed.), Philosophical writing of Peirce. New York: Dover Publications. 150-156.

Perrons, D. (2003a). The new economy and the work-life balance: conceptual explorations and a case study of new media. Gender, work & organisation, 10(1), 65-93.

Perrons, D. (2003b). The new economy, labour market inequalities and the work life balance. In: Martin, R. & Morrison, P.S. (eds.). Geographies of labour market inequality. Routledge, London, 129-148.

Perrons, D. (2005). Gender mainstreaming and gender equality in the new (market) economy: An analysis of contradictions. Social politics: international studies in gender, state & society, 12 (3), 389-411.

Perrons, D. (2007). Living and working patterns in the new knowledge economy: new opportunities and old social divisions in the case of new media and care-work. In Walby, S., Gottfried, H., Gottschall, K. & Osawa, M, (Eds.). Gendering the knowledge economy: comparative perspectives. Palgrave Macmillan, Basingstoke, UK, 188-206.

Perrons, D., McDowell, L., Fagan, C., Ray, K. & Ward, K. (2007). Gender, social class & work-life balance in the new economy. In Crompton, R., Lewis, S. & Lyonette, C. (Eds.). Women, men, work and family in Europe. Palgrave Macmillan, Basingstoke, UK, 133-151.

Perry, C. (1998). Processes of a case study methodology for postgraduate research in marketing, European Journal of Marketing, 32 (9/10), 785-802.

Pham, A. (2008, October 21). Women left on sidelines in video game revolution, Los Angeles Times, retrieved 20 August 2009.

Pini, M. (2001). Club Cultures and Female Subjectivity: The Move from Home to House, London: Palgrave.

Pocock, B. (1998). All Change, Still Gendered : the Australian Labour Market in the 1990s, Journal of Industrial Relations, 40, 580-604.

Poole, M.S. & Van de Ven, A.H. (1989). Using paradox to build management and organization theories. Academy of Management Review, 14, 562–578.

Pozzebon, M. & Pinsonneault, A. (2005). Challenges in Conducting Empirical Work Using StructurationTheory: Learning from IT Research, Organization Studies, 26 (9), 13531376.

Prasad, A. (2002). The Contest over meaning: Hermeneutics as an Interpretive Methodology for Understanding Texts, Organizational Research Methods, 5, 12-33.

Pratt, A.C. (2000). New Media, the New Economy and New Spaces’, Geoforum, 31(4), 425–36.

Pratt, A.C. (2006). Advertising and Creativity, a Governance Approach: A Case Study of Creative Agencies in London, Environment and Planning, 38(10), 1883–99.

Pratt. M. (2007). Computer game industry looks to women for fresh insights. Retrieved 18 Sept 2010 fromhttp://www.computerworld.com/s/article/293317/Computer_game_industry_looks_to_women_for_fresh_insights

Pratt, A.C., Gill, R. & Spelthann, V. (2007). Work and the city in the E-society A critical investigation of the sociospatially situated character of economic production in the digital content industries in the UK, Information, Communication & Society, 10(6), 922-942.

Prescott, J. & Bogg, J. (2010). The Computer Games Industry: Women’s Experiences of Work Role in a Male Dominated Environment. In Cater-Steel, A. & Cater, E. (Eds.), Women in Engineering, Science and Technology: Education and Career Challenges, 138-158.

Prescott, J., & Bogg. (2011). Segregation in a Male-Dominated Industry: Women Working in the Computer Games Industry. International Journal of Gender, Science and Technology, Vol.3, No.1.

Puth, G. (1996). Dynamics of the Marketing Environment: Implications for Marketing Training in South Africa. Unpublished monograph. Department of Business Economics, University of Pretoria

Queensland Government. (2001). Beyond the Pink Collar: Towards Strategies to Respond to Women and Work Issues in Queensland. Premiers Council for Women Brisbane, Australia: Queensland Government.

Queensland Government. (2003). Women in the Smart State Directions Statement 2003–2008. Brisbane, Australia: Queensland State Government. Retrieved October 19, 2006 from http://www.communities.qld.gov.au/resources/women/resources/directions-statement/directions-statement-text.pdf

Queensland Government. (2006a). Issues paper from the National ICT skills summit, June 21 & 22, Brisbane, Australia: Queensland State Government. Retrieved 10Oct 2010 from http://203.210.126.185/dsdweb/v4/apps/web/secure/docs/2406.pdf

Queensland Government. (2006b). Smart Women – Smart State Taskforce Women’s Participation in Emerging Industries: Science, Engineering and Technology Concept Paper. June 2006. Supporting Women’s Participation in Emerging Industries: Science, Engineering and Technology, Brisbane, Australia: Queensland State Government, Retrieved Oct, 2011 from http://www.communities.qld.gov.au/resources/women/work-and-life/set-concept-paper.pdf

Queensland Government. (June 2006c). Smart Women – Smart State Strategy Science, Engineering and Technology Action Plan 2006 – 2009. Queensland Government Office for Women. Brisbane, Australia: Queensland Government. Retrieved June 2007 from http://www.communities.qld.gov.au/resources/women/work-and-life/set-action-plan.pdf

Quesenberry, J. L. (2006). Career Anchors and Organizational Culture: A Study of Women in the IT Workforce. Computers and Society, 342-344.

Quesenberry, J., Trauth, E.M. & Morgan, A. (2006). Understanding the ‘Mommy Tracks’: A Framework for Analyzing Work-Family Issues in the IT Workforce, Information Resources Management Journal 19(2), 37-53

Raduescu, C. & Vessey, I. (2009). Methodology in Critical Realist Research: The Mediating Role of Domain Specific Theory. AMCIS 2009 Proceedings. Paper 433. http://aisel.aisnet.org/amcis2009/433/

Ragins, B.R., & Cotton, J.L. (1991). Easier said than done: Gender differences in perceived barriers to gaining a mentor. Academy of Management Journal, 34, 939-951.

Ramsey, N. & McCorduck, P. (2005). Where are the Women in Information Technology? Report of Literature Search and Interviews. Prepared by the Anita Borg Institute for Women and Technology (www.anitaborg.org) for the National Center for Women & Information Technology. National Science Foundation under Grant No. 0413538. Retrieved August 2nd, 2006, from http://www.anitaborg.org/news/publications/wherearethe_women.pdf

Reed, M.I. (1997). In praise of duality and dualism: rethinking agency and structure in organizational analysis. Organization Studies, 18(1), 21-42.

Reed, M.I. (2001). Organization, trust and control: a realistic analysis. Organization Studies, 22(2), 210-228.

Reskin, B. (2003). Including Mechanisms in Our Models of Ascriptive Inequality: 2002 Presidential Address. American Sociological Review 68(February), 1-21. Retrieved from http://www.asanet.org/images/members/docs/pdf/featured/Reskin_2003.pdf

Rettenmayer, J., Berry, R. & Ellis, S. (2007). The Image Of The Information System Profession: An Investigation Of Gender Differences. Journal of Computing Sciences in Colleges, 22 (5), 46-51.

Richards, L. (2005). Handling qualitative data. London: Sage.

Risman, B. J. (2004). Gender as a social structure: Theory wrestling with activism. Gender and Society, 18(4), 429-450.

Ritzer. G. & Smart, G. (Eds). (2001). Handbook of Social Theory. London: Sage.

Roan, A. & Whitehouse, G. (2007). Women, Information Technology and Waves of Optimism: Australian Evidence on Mixed-Skill Jobs. In New Technology, Work and Employment, 22(1), 21-33, March 2007. Available at SSRN: http://ssrn.com/abstract=968040 or doi:10.1111/j.1468-005X.2007.00181.x

Roldan, M., Soe, L. & Yakura, E.K. (2004). Perceptions of chilly IT organizational contexts and their effect on the retention and promotion of women in IT. In Proceedings of the 2004 SIGMIS Conference on Computer Personnel Research: Careers, Culture, and Ethics in A Networked Environment (Tucson, AZ, USA, April 22 - 24, 2004). SIGMIS CPR '04. ACM Press, New York, NY, 108-113. DOI= http://doi.acm.org/10.1145/982372.982399

Roodhouse, S. (2006). The Creative Industries: Definitions, Quantification and Practice, Conference Proceedings Cultural Industries: The British Experience in International Perspective Cultural Industries, Centre for British Studies, Humboldt-Universität, Berlin, http://www.62bug.net/roodhouse/docs/Humboldt_Univ%20paper06CI.pdf>last accessed 20 Aug 2007.

Roos, A.P & Gatta, M.L. (2009). Gender (in).equity in the academy: Subtle mechanisms and the production of inequality, Research in Social Stratification and Mobility, 27(3), 177-200. Retrieved http://www.sciencedirect.com/science/article/B82Y4-4W7B0K0-2/2/452c69f50fe8159525758ad1c703395b).

Rose, J. (1998). Evaluating The Contribution Of Structuration Theory To The Information Systems Discipline. http://folk.uio.no/patrickr/refdoc/ECIS1998-Rose.pdf

Rose, J. & Scheepers, R. (2001). Structuration theory and information system development frameworks for practice, ECIS 2001: Global Co-operation in the New Millennium, Bled, Slovenia, 217-231.

Rowlands B. (2003). Employing Interpretive Case Research to Build Theory of Information Systems Practice, Australian Journal of Information Systems, 10(2), 3-22.

Rowlands B. (2005). Grounded in Practice: Using Interpretive Research to Build Theory. The Electronic Journal of Business Research Methodology, 3(1), 81-92 Retrieved 19September 2006 from http://www.ejbrm.com/vol3/v3-i1/v3-i1-art7-rowlands.pdf

Sackmann, S.A. (1992). Culture and subcultures: An analysis of organizational knowledge. Administrative Science Quarterly, 37, 140–161.

Säljö, R. (2004). Learning and technologies, people and tools in co-ordinated activities. International Journal of Educational Research, 41, 489–494.

Saunders, P. (2002). Mutual Obligation, Participation and Popularity: Social Security Reform in Australia, Journal of Social Policy, 31(1), 21-38.

Savickas, M.L. (1997). Career adaptability: An integrative construct for life-span, life-space theory. The Career Development Quarterly, 45, 247-259.

Savickas, M.L. (2005). The theory and practice of career construction. In Brown, S. D. & Lent, R.W. (Eds). (2005). Career development and counseling: Putting theory and research to work. New York: Wiley.

Sayer, A. (1992). Method in Social Science – A Realist Approach, 2nd ed. London: Routledge.

Sayer, A. (2008). Case studies and a single, differentiated reality: A reply to Sue Llewellyn (2007), Qualitative Research in Accounting & Management, 5(1), 11 – 14.

Sayer. A. (2000). Realism and Social Science, Sage, London.

Scanlan, C.L., & Darkenwald, G. (1984). Identifying deterrents to participation in continuing education. Adult Education Quarterly 34(3), 155-166.

Schein, E.H. (1985). Organizational culture and leadership: A dynamic view. San Francisco, CA: Jossey-Bass.

Schön, D.A. (1983). The Reflective Practitioner. How professionals think in action, London: Temple Smith.

Seligsohn, I.J. (1967). Your Career in Computer Programming. Julian Messner.

Selwyn, N. (2002). Defining the ‘Digital Divide’: Developing a Theoretical Understanding of Inequalities in the Information Age. Occasional Paper 49. Cardiff: Cardiff University. Retrieved August 16, 2006 from http://www.cf.ac.uk/socsi/ict/definingdigitaldivide.pdf

Sen, A. (2000). Social Exclusion: Concept, Applications, and Scrutiny, Social Development Papers No. 1, Office of Environment and Social Development, Asian Development Bank.

Servon, L. & Nelson, M. (2001).Community Technology Centres: Narrowing the digital divide in low-income, urban communities. Journal of Urban Affairs, 23(3-4), 279-290.

Sethibe, T., Campbell, J. & McDonald, C. (2007). IT Governance in Public and Private Sector Organisations: Examining the Differences and Defining Future Research Directions. 18th Australasian Conference on Information System 5-7 Dec 2007, Toowoomba, Australia.

Sewell, W.H. (1992). A Theory of Structure: Duality, Agency, and Transformation. The American Journal of Sociology, 98(1), (July, 1992), 1-29.

Shinkle, E. (2008). Video games, emotion and the six senses. Media Culture Society, 30, 907-915.

Shoib, G., Nandhakumar, J. & Jones, M. (2006). Using social theory in information systems research: A reflexive account, in quality and impact of qualitative research. Proceedings of QualIT2006: Quality and Impact of Qualitative Research, Brisbane, Australia, Griffith University.

Sicart, M. (2009). The Ethics of Computer Games. Cambridge: The Massachusetts Institute of Technology Press.

Silverman, D. (2009). Doing qualitative research (3rd ed.). London: Sage.

Silverstone, R. (1999). Why Study The Media? London: Sage.

Skillset (2004). Labour Force Employment survey. Retrieved June 2012 from http://www.creativeskillset.org/uploads/pdf/asset_9920.pdf

Smith, D.H. (1980). Participation in Social and Political Activities, Smith, D.H., Macaulay, J. & Associates (Eds). San Francisco: Jossey Bass.

Smith, M. (2006). Overcoming theory-practice inconsistencies: Critical Realism and information systems research. Information and Organization, 16(3), 191-211.

Smith, M.J. & Liehr, P.R. (2008). Middle range theory for nursing. New York: Springer Publishing.

Smith, S.M. (2002). Using the Social Cognitive Model to Explain Vocational Interest in Information Technology. Information Technology, Learning, and Performance Journal. Spring, 20(1), 1-9.

Snow, D. & Anderson, L. (1991). Researching the Homeless: The Characteristic Features of the Case Study in Feagin, J., Orum, A., & Sjoberg, G. (Eds.). A case for case study. Chapel Hill, NC: University of North Carolina Press. 148-173.

Spender, D. (1995). Nattering on the Net: women power and cyberspace. Melbourne, Spinifex.

Spiers, J.A. (2004). Tech tips: Using video management/ analysis technology in qualitative research. International Journal of Qualitative Methods, 3(1). Article 5. Retrieved 12/10/2007 from http://www.ualberta.ca/~iiqm/backissues/3_1/pdf/spiersvideo.pdf

Spradley, J.P. (1979). The Ethnographic Interview. Fort Worth: Harcourt Brace.

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Stake, R. E. (2000). Case studies. In Denzin, N. & Lincoln, Y. (eds.). Handbook of qualitative research (2nd ed). 435-454. California: Sage Publications.

Stangor, C., Carr, C. & Kiang, L. (1998). Activating stereotypes undermines task performance, Journal of Personality and Social Psychology, 75, 1191-1197.

State of Victoria. (2007). ICT skills research. Attitudes to ICT careers and study among 14-19 year old Victorians (Years 9-12). Melbourne, Vic: Multimedia Victoria.

Stockdale, R. & Stoney, S. (2008). Generating a Gender Balance: making introductory information systems courses a positive experience. Australasian Journal of Information Systems, 15(1). Retrieved 12 November 2011 from http://dl.acs.org.au/index.php/ajis/article/view/477

Stones, R. (2005). Structuration Theory, Basingstoke, UK: Palgrave-Macmilla

Stones, R. (2001). Refusing the Realism–Structuration Divide. European Journal of Social Theory 4(2).: 177–197

Straub, D., Loch, K., Evaristo, R., Karahanna, E. & Srite, M. (2002). Toward a theory-based measurement of culture. Journal of Global Information Management, 10(1), 13-23.

Strauss, A., Schatzmann, L., Bucher, R., Ehrlich, D. & Sabshin, M. (1964). Psychiatric Ideologies and Institutions. Glencoe Ill. The Free Press.

Summer, M. & Niederman. F. (2002). The impact of gender differences on job satisfaction, job turnover, and career experiences of information systems professionals, in Proceedings of the 2002 ACM SIGCPR Conference, Kristiansand, Norway, 154-163.

Super, D.E. (1992). Toward a comprehensive theory of career development. In Montross, D. & Shinkman, C. (eds.), Career development: theory and practice. 35–64. Springfield,IL: Charles C. Thomas.

Sutton R.I. & Staw, B.M. (1995). What Theory is not. Administrative Science Quarterly, 40(3), 371-38

Swanson, G & Wise, P (1996). Digital Futures: Women’s Employment in the Multimedia Industries, Canberra: Commonwealth Government of Australia. Women's Research and Employment Initiatives Program (Australia). Department of Employment, Education, Training and Youth Affairs, Canberra : www.workplace.gov.au/NR/rdonlyres/CC7EFA3F-1E9F-4E77-ADC1 1A97C884CADB/0/digitalfutures.PDF

Swanson, J. L., and Woitke, M.B. (1997) Theory into practice in career assessment for women: assessment and interventions regarding perceived career barriers. Journal of career assessment, Vol. 5, pp.443-462.

Swartz, D.L. & Zolberg, V.L. (Eds). (2004). After Bourdieu. Influence, Critique, Elaboration, Dordrecht, Boston, London: Kluwer Academic Publishers.

Takahashi, D. (2010). Ethics Of Game Design. Gamsutra. Retrieved. Aug 2010 from http://www.gamasutra.com/view/feature/2181/ethics_of_game_design.php

Tapia, A., (2006). Hostile Work Environment.Com: Increasing Participation of Underrepresented Groups, Lessons Learned from the Dot-Com Era. Database for Information Systems Research. 37(4), 79-98.

Tapia, A., & Kvasny, L. (2004). Recruitment is never enough: retention of women and minorities in the IT workplace. Proceedings of the ACM Conference on Computer Personnel Research: Careers, Culture, and Ethics in a Networked Environment, Tucson, 84-91.

Tapia, A., Kvasny, L. & Trauth, E. (2004). Is There an IS Retention Gap for Women and Minorities? The Case for Moving In versus Moving Up, In Igbaria, M. and Shayo, C. (Eds.), Strategies for Managing IS/IT Personnel, Idea Group Publishing, Hershey: Idea Group, 143-164.

Tashakkori, A., & Teddlie, C. (Eds.). (2003). Handbook of mixed methods in social and behavioral research. Thousand Oaks, CA: Sage.

Taylor, T.L (2008). Becoming a Player: Networks, Structures, and Imagined Futures, in Kafai, Y. Heeter, C. Denner, J. and Sun, J. (Eds.). Beyond Barbie and Mortal Kombat: New Perspectives on Gender, Games, and Computing, Cambridge: The MIT Press.

Taylor, C. & White, S. (2000). Practising reflexivity in health and welfare: Making knowledge. Buckingham: Open University Press.

Teague, J. (1997). Structured Review of Reasons For the Under Representation of Women in Computing, Proceedings of the Second Australasian Conference on Computer Science Education, University of Melbourne, Australia, July 2-4.

Thomas, R. & Allen, A. (2006). Gender differences in students’ perceptions of information technology as a career. Journal of Information Technology Education, 5(5), 165-178.

Thompson, J. (1984). Studies in the theory of ideology. Cambridge, UK: Polity.

Thomson, A.J., & Schmoldt D.L. (2001). Ethics in computer software design and development. Computers and Electronics in Agriculture, 30 (2001) 85–102

Thrift, N. & May, J. (Eds.). (2001). Timespace: Geographies of Temporality. London: Routledge

Timms, C., Courtney, L. & Anderson, N. (2006). Secondary girl perceptions of advanced ICT subjects: Are they boring and irrelevant? Australian Educational Computing Research News, 21(2), 3-8.

Trauth, E.M. (1999). Who Owns my Soul? The Paradox of Organizational Productivity in a Culture of Individualism. Proceedings of Association for Computing Machinery Special Interest Group on Computer Personnel Research Annual Conference: Managing Organizational Knowledge for Strategic Advantage - The Key Role of Information Technology and Personnel (New Orleans, L.A.). 159-163.

Trauth, E.M. (2001). Qualitative Research in IS: Issues and Trends, Idea Group Publishing, Hershey, PA.

Trauth, E.M. (2002). Odd Girl Out: An Individual Differences Perspective on Women in the IT Profession, Information Technology and People, 15(2), 98-118.

Trauth, E.M. (2006). Theorizing Gender and Information Technology Research, in Encyclopaedia of Gender and Information Technology. Trauth, E.M. (Ed.), Hershey, PA: Idea Group Publishing: 1154-1159.

Trauth, E.M. (2011).What Can We Learn from Gender Research? Seven Lessons for Business Research Methods. The Electronic Journal of Business Research Methods. 9(1), 1-9, available online at www.ejbrm.com

Trauth, E.M., & Howcroft, D. (2006). Critical Empirical Research in IS: An Example of Gender and IT, Information Technology and People. Special Issue on Critical Research in Information Systems 19(3) 272-292.

Trauth, E.M., Joshi, K.D., Kvasny, L., Chong, J., Kulturel, S. & Mahar. J. (2010). Millennials and Masculinity: A Shifting Tide of Gender Typing ICT? Proceedings of the 16th Americas Conference on Information Systems (Lima, Peru, August).

Trauth, E.M., Nielsen S.H., & von Hellens, L.A. (2003). Explaining the IT gender gap: Australian stories for the new millennium. Journal of research and practice in information technology, 35(1), 7-20. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library database.

Trauth, E.M., & Quesenberry, J.L. (2005). Individual inequality: women's responses in the IT profession, paper presented at the Women, Work and IT Forum.

Trauth, E.M., & Quesenberry, J.L. (2007). Gender and the information technology workforce: issues of theory and practice, in Young, P., Huff, S. (Eds), Managing IT Professionals in the Internet, Idea Group Publishing, Hershey, PA, 18-36.

Trauth, E.M., Quesenberry, J.L. & Huang, H. (2008). A Multicultural Analysis of Factors Influencing Career Choice for Women in the Information Technology Workforce, Journal of Global Information Management, 16(4), 1-23.

Trauth, E.M., Quesenberry, J. & Huang, H. (2009). Retaining women in the U.S. IT workforce: theorizing the influence of organizational factors. European Journal of Information Systems. 18(5), 476-497. http://dx.doi.org/10.1057/ejis.2009.31

Trauth, E.M., Quesenberry, J. & Morgan, A.J. (2004). Understanding the Under Representation of Women in IT: Toward a Theory of Individual Differences. Proceedings of the ACM SIGMIS Computer Personnel Research Conference: Careers, culture, and ethics in a networked environment, SIGMIS'04, April 22-24, 2004. 114-119.Tucson. AZ. ACM Press: USA. DOI=10.1145/982372.982400 http://doi.acm.org/10.1145/982372.982400 Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

Trauth, E.M., Quesenberry, J. & Yeo, B. (2005). The influence of environmental context on women in the IT workforce. Proceedings of the 2005 ACM SIGMIS CPR conference on Computer personnel research. April 14-16, 2005, Atlanta, Georgia, USA, 24 – 31. ACM Press: USA. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

Trauth, E. M., Quesenberry, J. L. & Yeo, B. (2008b). Environmental influences on gender in the IT workforce. ACM SIGMIS Database, 39(1), 8-32.

Tsoukas, H. (1989). The validity of idiographic research explanations. Academy of Management Review, 14(4), 551-561.

Turkle, S. (1995). Life on the screen: Identity in the age of Internet. New York: Simon & Schuster.

Turner, J.H. (1991) Structuration Theory of Anthony Giddens. In The Structure of Sociological Theory (5thed). California: Wadsworth, pp. 519 – 539.

Turner, S. V., Bernt, P. W. & Pecora, N. (2002). Why women choose information technology careers: Educational, social and familial influences. Proceedings of the annual meeting of the American Educational Research Association, New Orleans, LA.

Turner, T. (2003). Multimedia is much more fun but I still don’t want to do IT. Proceedings of the 2003 Australian Women in IT Conference. Retrieved September 19, 2006 from http://www.auswit.org/2003/papers/3-2_Turner.pdf

UNCTAD (United Nations Committee on Trade, Aid and Development). (2008). Creative economy report. Geneva: UNCTAD.

Valian, V. (1998). Why so slow?: The advancement of women. Cambridge: The MIT Press.

Van Manen, (1990). Researching Lived Experience: Human Science for an Action-Sensitive Pedagogy (London, Ontario: The University of Western Ontario, 1990), 175.

Varella, C. & Harre, R. (1996). Conicting varieties of realism: Causal powers and the problems of social structure, Journal for the Theory of Social Behaviour, 26(3).

Venkatesh, V. & Morris, M.G. (2000). Why Don’t Men Ever Stop to Ask for Directions? Gender, Social Influence, and Their Role in Technology Acceptance and Usage Behavior? MIS Quarterly, 24(1), 115-139.

Verba, S., & Nie, N.H. (1972). Participation in America. New York: Harper & Row.

Vigdor, L. (2011). A Techno-Passion that is Not One: Rethinking Marginality, Exclusion and Difference. International Journal of Gender, Science and Technology, 3(1).

Volkoff, O., Strong, D.M. & Elmes, M.B. (2007). Technological Embeddedness and Organizational Change, Organization Science, 18(5), 832-848.

von Hellens, L., Clayton, K., Beekhuyzen, J., & Nielsen, S. (2009). Perceptions of ICT careers in German schools: an exploratory study. Journal of Information Technology Education, 8, 211-228.

von Hellens, L., Nielsen, S.H. & Beekhuyzen, J.P. (2004). An exploration of the dualisms in the female perceptions of IT work. Journal of Information Technology Education, 3, 103-116. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

von Hellens, L., Nielsen, S.H. & Beekhuyzen, J. (2003). Women working in the IT industry: Challenges for the new millennium. International Business and Economics Research Journal, 1(11).

von Hellens, L., Nielsen, S., & Trauth, E.M. (2001). Breaking and entering the male domain: Women in the IT industry. In Proceedings of the 2001 ACM SIGCPR Computer Personnel Research Conference.

von Hellens. L., & Nielsen. S. (2001). Australian women in IT. Communications of the ACM, 44(7), 46-52. Retrieved August 2nd, 2005, from ACM Portal from The ACM Digital Library Database.

Wajcman, J. (1991). Feminism Confronts Technology. Proceedings from 'In Action for Equity: The Second Decade'. The Sixth international GASAT conference, Vol. 2. Allen and Unwin, Sydney.

Wajcman, J. (2000). Reflections on gender and technology studies: in what state is the art?’Social Studies of Science, (30)3, pp. 447-64

Walsham, G. (1993). Interpreting Information Systems in Organisations, Wiley, Chichester.

Walsham, G. (1995a). The emergence of interpretivism in IS research. Information Systems Research, 6(4), 376-394.

Walsham, G. (1995b). Interpretive case studies in IS research: nature and method, European Journal of Information Systems (4), 74-81.

Walsham, G. (2002). Cross-cultural software production and use: A structurational analysis. MIS Quarterly, 26(4): 359-380.

Walsham, G. (2005). Learning about Being Critical. Information Systems Journal, 15, 111-117.

Walsham, G. (2006a). Doing interpretive research. Summary for seminar. Retrieved August 2010 http://www.dipartimenti.unicatt.it/segesta_Walsham.pdf

Walsham, G. (2006b). Doing interpretive research, European Journal of Information Systems, 15, 320-30.

Walters, K. (2006). EXIT: Why Women are Shunning the Technology Industry, Business Review Weekly, July, 26-32.

Walton, M. & Duncan, I. (2002). Creative Industries in New Zealand Economic Contribution, Industry New Zealand, Wellington, http://www.nzte.govt.nz/ common/files/nzier-mapping-ci.doc last accessed 21/08/2007.

Warne, L., Bandias, S., and Fuller, D. (2011). The Employment Experiences of Women in the Australian Information Communication Technology Industry. Journal of Economic and Social Policy, 14(1), Article 2. Available at: http://epubs.scu.edu.au/jesp/vol14/iss1/2

Weaver, G. and Gioia D. (1994). Paradigms Lost: Incommensurability vs Structurationist Inquiry, Organization Studies,15(4), 565-590.

Webb, P., and Young, J. (2005). Perhaps it’s time for a fresh approach to ICT gender research? [Electronic version] Journal of Research and Practice in Information Technology, 37(2), 147-160. Retrieved September 17, 2006 from, http://www.jrpit.acs.org.au/jrpit/JRPITVolumes/JRPIT37/JRPIT37.2.147.pdf

Weber, R. (2003a). Editors Comments: The reflexive researcher, MIS Quarterly 27, (4). September, v-xiv.

Weber, R. (2003b). Editors Comments: theoretically Speaking, MIS Quarterly 27(3). September, iii-xiii.

Webster, M., Jr., and Whitmeyer, J.M. (1999). A Theory of Second-Order Expectations and Behavior. Social Psychology Quarterly 61, 17-31.

Weick, K. E. (1989). Theory construction as disciplined imagination. Academy of Management Review, 14(4), 516-531.

Weick, K.E. (1995). What theory is not, theorizing is, Administrative Sciences Quarterly, 40(3), 385-390.

Weisinger, J. and Trauth, E. (2003). The Importance of Situating Culture in Cross-Cultural IT Management, IEEE Transactions on Engineering Management, 50(1), 26-30.

Wenger, E. (1998). Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cambridge University Press.

Wenger, E, McDermott, R.,Snyder, W. M. (2002). Cultivating Communities of Practice (Hardcover). Harvard Business Press; 1 edition. ISBN 978-1-57851-330-7.

Wengraf, T. (2001). Qualitative Research Interviewing: Biographic Narrative and Semi-Structured Methods, London, Sage.

Whetten, D.A. (1989). What constitutes a theoretical contribution? Academy of Management Review, 14(4), 490-495.

Whitehouse, G., and Diamond, C. (2005). Reproducing gender inequality: Segregation and career paths in information technology jobs in Australia. Keynote address to the 19th Annual Conference of the Association of Industrial Relations Academics of Australia and New Zealand (AIRAANZ), 10 February 2005, University of Sydney, Sydney.

Wikgren, M. (2005). Critical Realism as a philosophy and social theory in information science? Journal of Documentation, 61(1), 11-22.

Williams, R. (1983). Keywords: A Vocabulary of Culture and Society. London: Fontana

Willmott, R., (1997). Structure, Culture and Agency: Rejecting the Current Orthodoxy of Organisation Theory. Journal for the Theory of Social Behaviour, 27(1), 0021-8308, Blackwell, Oxford.

Wilson, M. (2004). A conceptual framework for studying gender in information systems research. Journal of Information Technology, 19, 81-92.

WIT (2006). Women in Information Technology - WIT Home Page/Mission. Accessed September 17, 2006 from, http://www.wit.org.au

Witkowski. E. (2006). Soft Reboot: Young Women and Technology in Denmark. Independent research for IBM Denmark.

Witzel, A. (2000, January). The problem-centered interview. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [On-line Journal], 1(1). Available at: http://qualitative-research.net/fqs

Women on the Web Report (2010). How Women are Shaping the Internet. Reston, VA Retrieved Aug 2010 http://mashable.com/2010/07/28/women-on-the-web/

Woolley, H.T. (1910). A Review of the Recent Literature on the Psychology of Sex, Psychological Bulletin 7 (18), 335-42.

Yeung, H.W. C. (1997).Critical Realism and realist research in human geography: A method or a philosophy in search of a method? Progress in Human Geography, 21(1), 51 74.

Yin, R. (1984). Case study research: Design and methods (1st Edition). Beverly Hills, CA: Sage Publishing.

Yin, R. (1993). Applications of case study research. Beverly Hills, CA: Sage Publishing.

Yin, R. (1994). Case study research: Design and methods (2nd ed.). Beverly Hills, CA: Sage.

Yin, R. (2003). Case study research: Design and methods (3rd ed., Vol. 5). Thousand Oaks, CA: Sage.

Yin, R., (2004). The Case Study Anthology, Sage, Thousand Oaks, CA

Yin, R (2011). Applications of Case Study Research. Third Edition. Sage.

Yow, V. R. (1994). Recording oral history: A practical guide for social scientists. Thousand Oaks, CA: Sage.

Zheng, Y. & Walsham, G. (2008). Inequality of what? Social exclusion in the e-society as capability deprivation, Information Technology and People, 21(3), 222-43.

Appendices ‘Women’s Participation in the Australian Digital Content Industry’Author: Anitza Geneve Year: 2013 Page 350 of 401

Appendices

APPENDIX 1: PUBLICATIONS STEMMING FROM THE RESEARCH

Type Publication Date/ Country

Conference presentation

Geneve, A., (2011). “Five Acts of Agency by women in the Australian Digital Content Industry”. Console-ing Passions conference, 27-29 July, Adelaide, Australia 2011

2011 Australia

Book Chapter Geneve, A., and Ganito, C. (2011) “Women and Technology: Five Acts of Acts of Agency" in Araya, D. Breindl, Y, Houghton, T (Ed.) Nexus: New Intersections in Internet Research. Peter Lang:USA ISBN 978-1-4331-0970-6 pb.

2011 USA

Conference panel

Ganito, C., Breindl,Y., Haché, A., Geneve, A., Blodgett,B., Ferreira, C., Lopez, E., Vergés Bosch, N., (2010) Panel: Quantity and Quality: Patterns of Female Engagement with ICTs. EquityInternet Research 11, October 23-24, 2010,

2010 Sweden

Book Chapter Geneve, A., and Christie, R.J. and Nelson, K. J. (2009) Women's participation in the Australian Digital Content Industry: initial case study findings. In: Prpi , K and Oliveira, Laura and Hemlin, S (Eds.) Women in Science & Technology. Institute of Social Research-Zagreb Sociology of Science and Technology Network of ESA (European Sociological Association), Zagreb, Croatia, pp. 139-161. ISBN: 9789536218417

2009 Croatia

Conference presentation

Geneve, A., (2011). Sphere of Influence and 'Five Acts' of 'Acts of Agency'. OzWIT 2009: Australian Celebration of Women in ICT, Melbourne, Australia.

2009 Australia

Conference paper

Geneve, A., and Nelson, K.J. and Christie, R. J. (2008) Passion, women and the games industry: influences on women's participation in the Australian Digital Content Industry. In: Women in Games, 10-12th September 2008, Warwick, United Kingdom. (Unpublished)

Sept 2008 United Kingdom

Conference paper

Geneve, A., Nelson, K., and Christie, R. (2008). Girly Girls and Geeky Guys. Pathways of desire: Women’s participation in the Digital Content Industry. Investigating the theory and practice of gender and sexuality in the workplace. 12 - 13 June 2008. The Institute of Advanced Studies, University of Surrey, United Kingdom.

June 2008 United Kingdom

Conference paper

Geneve, A., Nelson, K., and Christie, R. (2008). How did ‘we’ get here? Pathways of desire: Women’s participation in the Digital Content Industry. The European Sociological Association SSTNET (Sociology of Science and Technology Network) Workshop, 5 - 6 June 2008, Zagreb, Croatia.

June 2008 Croatia

Doctoral consortium paper

Geneve, A. (2005). Does confidence assist women in responding to factors that influence their participation in Information Communication Technology (ICT)? 2nd International Conference on Qualitative Research in IT & IT in Qualitative Research- Doctoral Consortium. November 23 - 25, 2005, Griffith University Brisbane, Australia.

Nov 2005 Australia

* peer-reviewed

APPENDIX 2: EXAMPLE OF LITERATURE INDICATING WOMEN ARE UNDER-REPRESENTED IN THE DCI AND RELATED ICT INDUSTRY IN AUSTRALIA AND INTERNATIONALLY

Evidence Country ICT DCI Key points

Special interest group (SIG) IGDA (current) http://archives.igda.org/women/

USA x Aims to address the under-representation of women in the games industry.

IGDA Diversity report (Gourdin, 2005)

x Identifies women are under-represented in the games industry internationally, although most respondents were from the USA.

Communications of the ACM (1990)

x Suggests many women opting for careers in computing drop out of the academic pipeline.

Communications of the ACM (1995)

x Special issue on women in computing features a historical overview as well as surveys of conditions in various countries.

Communications of the ACM ACM (2002)

x Propose recommendations to recruit and retain women in computing at the undergraduate level, including mentoring and role models.

ACM SIGCSE Bulletin (2002)

x Identifies the pipeline shrinkage problem for women in computer science as being a well-known and documented phenomenon.

Women in the Smart State Directions Statement 2003-2008 (Queensland Government, 2003)

QLD Australia

x x Not limited to ICT but does not explicitly mention DCI.

Queensland Government. (2006b)

QLD Australia

x x Makes reference to “supporting women’s participation in emerging industries” (Queensland Government, 2006b, p.1) but does not explicitly mention DCI.

partICipaTion Summit (Queensland Government, 2006a)

Australia x Identified a “need for the industry to ensure […] work environment is attractive and flexible to attract and retain people in the ICT profession”.

AAUW, 2004 USA x Outlines the increased government funding in the United States towards increasing the number of girls and women participating in science, technology, engineering and math.

Queensland Government (2001)Beyond the Pink Collar: (Premiers Council for Women)

QLD Australia

x x Profile of women’s positions in the Queensland workforce and an analysis of the impact of key Government policy and initiatives on work-related outcomes for women.

APPENDIX 3: EXAMPLES OF STRATEGIES FOSTERING WOMEN PARTICIPATION

Table 55 Examples of strategies fostering girls and women’s participation Strategy /Initiative

Industry1 Area2 Country Example of type of activity

Typical cohort

WINIT http://www.winitproject.com/index.html

ICT Education UK, AU Mentoring Student university

Digital Divas http://www.digitaldivas.com.au/

ICT Education AU Curriculum, role models,

Students aged 12-15

WIT http://www.wit.org.au/

ICT/DCI Industry, Government

UK, AU female role models, networking,

Board readiness programs, mentoring programs for women in STEM industries

IGDA mentorship scheme http://www.igda.org/scholars/

DCI (Games)

Industry USA 3 Mentoring Online mentoring for games workers with a focus on women, now a scholarship program

Sun SEED mentoring program http://www.facebook.com/pages/SEED-Engineering-Mentoring-Program/165811522614

ICT Education, Industry

AU Mentoring, opportunity for experience

Industry based work experience, now a Facebook community

Technology Takes you Anywhere https://www.learningplace.com.au/defaulteqa2.asp?orgid=48&suborgid=665

ICT/DCI Education AU role models, hands on experience

Once per year visit by school aged girls to event(QLD)

Diploma of Multimedia (Women) 4 DCI Education AU Skills development

Training

Girl Geek Coffee Club https://sites.google.com/site/girlgeekcoffees/

ICT/DCI Education AU Networking Networking for female university students

Geek Girl Dinners http://girlgeekdinners.com/

Education, Industry

UK 3 Networking Networking focus for female students, industry workers

Go for IT gURL http://www.qut.edu.au/study/fees-and-scholarships/scholarships-and-prizes/go-for-it-gurl-merit-scholarship

ICT Education AU Role models, information

Once per year visit to university (QUT), late school aged now merit based scholarship

Go girl go for IT http://gogirl.org.au/

ICT Education AU School to university pathway

Once per year visit by school aged girls to event (Victoria)

Women in Games http://www.womeningames.com/

DCI (Games)

Industry USA Networking, information for games workers

Online website offering resources such as social networking

WITI http://www.witi.com/

ICT Industry AU3 Networking Online website and in person meetings to WITI to help women advance by providing access to - and support from - other professional women working in all sectors of technology http://www.witi.com/

Google Anita Borg scholarship http://www.google.com.au/anitaborg/

ICT Education AU, USA

Networking, financial

Offers female recipients a $5,000 scholarship and visit to Googleplex in Sydney, Australia

FITT Females in Information Technology & Telecommunications http://www.fitt.org.au/

Promotes the interests of women working in the Information and Communications Technology (ICT) industry in Australia

Women in Games Australia http://www.womeningames.org.au/

Games Industry AU Information, networking

An organisation run for women in the gaming industry Last post nov 2011

1 Industries often overlap, 2 does not closely examine sources of funding 3 international strategy 4 last offered as a gender specific program in 2007

APPENDIX 4: INFLUENCES ON WOMEN’S PARTICIPATION AS IDENTIFIED IN PREVIOUS RESEARCH

Table 56 provides examples of influences on women’s participation (in both the DCI and

ICT industry) as identified from the literature. The meta analysis of literature was used to sensitise the

researcher and inform the development of Framework 1, an initial guiding analytical model (see

Section 3.1).

Table 56 Example of influences on participation as identified in ICT and DCI literature Influence Author Year *** Industry

sectorExample of influence Women

specific/cohort*/**

Culture-family Trauth, Quesenberry, Huang (2008)

2006 USA IT Culture may differ between countries

Organizational culture

Trauth, Quesenberry, Huang (2009)

2009 USA IT Y

Roldan, Soe, Yakura (2004), Webb and Young(2005) citing Cuny and Aspray (2000)

- USA AU

Chilly culture Competitive ‘aggressive’ work atmosphere

Trauth, Quesenberry and Yeo (2005)

2005 USA Attitudes, values regarding women, women working and women working in IT

Webb and Young (2005) citing von Hellens, Nielsen and Trauth (2001)

- AU Dominant male culture in IT organisations

Life course Castano and Webster (2011)

2011 *> Views influences over a life course perspective

Occupational stereotypes

Margolis and Fisher(2002); Jepson and Perl (2002); MacKnight (2001)

- UK USA

ICT (IT, CS) The media reinforces imagery of computer ‘nerds’ and ‘geeks‘

Stockdale and Stoney (2008), MacKnight (2001)

- - IT Stereotype of lonely, “nerdy” work …turns women off from seeking IT jobs.

Courtney, Lankshear, Anderson and Timms (2009)

2009 AU ICT Students and workers perceptions of a career in the IT industry differ

Gill (2002) UK/ EU

DCI (new media)

New Media seen as cool, creative egalitarian.

State of Victoria (2007) 2007 AU ICT/DCI (multimedia)

Multimedia seen as more fun than IT

Y High school students

Norms surrounding gender stereotypes

McGrath-Cohoon and Aspray (2006)

2006 USA ICT Female identity not compatible with computers for example, men presented as focused on computer and lacking social interests

Taylor (2008, p. 55) 2008 - The lack of women gamers is linked to the “image of what a “real women” is

Mentors Trauth et al., (2004); Trauth et al., (2009); Tapia and Kvasny (2004)

- USA ICT (IT) A lack of mentoring opportunities for women; women are more likely to have fewer mentors. Reluctance of male supervisors to coach and mentor women subordinates

Media Jepson and Perl (2002) 2002 - ICT (IT) “Negative media” convey the perception that computer-related careers are “too nerdy”

Y Education context, student cohort

Role models Jepson and Perl (2002) 2002 ICT(IT) Lack of female role models

Male majority Griffiths, Moore and Richardson (2007)

2007 UK ICT Masculinisation due to majority

Social networks

Morgan, Quesenberry & Trauth (2004)

2004 USA Networks are important

Table continued on next page

Influence Author Year *** Industry

sector Example of influence Women

specific/cohort*/**

Resource content

Klawe (2002) 2002 USA ICT Boy’s decision to pursue career linked to playing games, games content is targeted towards males

N School aged

Resource access (technology)

Adya and Kasier (2005) 2005 USA IT Access to technology in home and at school

Y Young girls

Long hours Gill (2002); Griffiths, Moore and Richardson (2007); Prescott and Bogg (2010)

- UK UK UK

DCI ICT DCI (games)

Long hours challenge life work balance

Y New media workers , ICT workers

Financial rewards

Panteli, Stack, & Ramsay, (1999)

1999 UK ICT Industry employment often contract based adds to employee insecurity, low pay

Available information

Jepson and Perl (2002) 2002 USA ICT Lack of information about certain career pathways

Tapia, Kvasny and Trauth in Igabaria and Shayo (2004)

2004 USA ICT insufficient information regarding role expectations

Rewards Joseph, Ng, Koh and Ang (2006)

2005 USA ICT Reward and advancement incentives are important motivators

N IT professionals

Economy Trauth, Quesenberry and Yeo (2008)

2008 USA ICT Size of the information economy, household income and cost of living

Geography Trauth, Quesenberry and Yeo (2008)

2008 USA ICT Geographical regions may offer different characteristics

Media Margolis and Fisher (2002) 2002 UK/ EU

Magazines Yes University students

Identity Moore, Griffiths and Richardson (2005)

2005 UK /EU

IT Identity tensions between women and techical

Personality traits

Trauth (2002), Trauth, Quesenberry, Yeo (2005); Trauth, Quesenberry, Yeo (2008)

2002 USA Confidence in women fosters participation, characteristics such as beign strong; internatl perseverance

Yes IT workers

Skills capacity Roan and Whitehouse, (2007)

2007 AU DCI Need for creative and not only technical skills

Yes DCI workers

Margolis and Fisher (2002) 2002 UK/ EU

IT Women face an ‘experience gap’ in regards to computer/technology use

Y University student

Self -efficacy Hackett (1995) 1995 ICT (IT) Self-efficacy fosters participation

Yes University students (male and female)

Psychological perceptions

Gilligan(1982) 1982 - Psychological differences between males and females

Need for social interaction

Misic & Graf, (1999) 1999 - DCI Limited opportunities for social interaction

Pratt (2000) 2000 USA DCI Workers need for social interaction

Berg et al. 2002 in McGrath-Cohoon and Asprey(2006)

2002 USA ICT Women value jobs they believe had social interaction and compatibility with social issues

Y Engineering students

Work-life balance

Trauth, Quesenberry, Huang (2006); Griffiths and Moore (2010); Moore, A. K., Griffiths, M., & Richardson, H. (2005b).

- USA UK UK

IT Women faced challenge in maintaining a life work balance

Personal values

McGrath-Cohoon and Asprey(2006)

2006 USA An IT career was a “waste of intelligence” and women wanted careers that “made a difference”.

Natural ability to work with computers

Trauth, Quesenberry, Yeo(2008)

2008 USA Y

*Study appeared to focus on women; however, the research may not have clearly indicated this. ** Identifies cohort if empirical study or if a review specifically mentions a cohort for example school aged children, *** Country of origin (may however encompass global aspects)

APPENDIX 5: MODELS REGARDING WOMEN’S PARTICIPATION FROM THE ICT LITERATURE

Figure 33: Model of Girls Career Choices (Adya and Kaiser, 2005).

Figure 34: Career stage model (Ahuja, 2002)

Figure 35: Webb and Young’s (2005) major and sub-themes

Table 57 Teague’s (1997) application of the PRECEDE-PROCEED model

PRECEDE The first 5 phases Phase 1 Social Diagnosis The focus of this phase is to identify and evaluate the social

problems which impact the quality of life of a target population

Phase 2 Epidemiological Diagnosis

Identify behavioural and environmental factors related to the quality of life issues such as statistics. Based on these two phases develop goals or outcomes for the program identifying who, what, how and when.

Phase 3 Behavioural & Environmental Diagnosis

Determination of the importance and relative changeability of each behavioural cause. Also personal factors not those controlled by behaviour.

Phase 4 Education & Organizational Diagnosis

Three kinds of causes are identified - predisposing factors, enabling factors, and reinforcing factors. The critical element of this phase is the selection of the factors which if modified, will be most likely to result in behaviour change. Predisposing Factors - any characteristics of a person or population that motivates behaviour prior to the occurrence of that behaviour

knowledge beliefs values attitudes

Enablers - characteristic of the environment that facilitate action and any skill or resource required to attain specific behaviour

accessibility availability skills laws (local, state, federal)

Reinforces - rewards or punishments following or anticipated as a consequence of behaviour. They serve to strengthen the motivation for behaviour.

family peers Teacher.

Phase 5 Administrative & Policy Diagnosis

Administrative and organizational concerns which must be addresses prior to program implementation aimed at changing behaviour.

PROCEED The second 4 phases Phase 6 Implementation Phase 7 Process Evaluation Phase 8 Impact Evaluation Phase 9 Outcome Evaluation

Teague applied the first five stages of the model (Predisposing, Reinforcing and Enabling

Constructs in Educational/environmental Diagnosis and Evaluation, and Policy, Regulatory, and

Organisational Constructs in Educational and Environmental Development)

APPENDIX 6: APPROACH TO ESTABLISHING ‘TRUSTWORTHINESS’ OF FINDINGS

Table 58 Approach to establishing ‘trustworthiness’ of the research findings Trustworthiness criteria

Recommended technique As implemented in the investigation Discussed in Section

Credibility (or internal validity in Positivism)

Prolonged engagement - Phase 1 data transcribed by researcher and verified by researcher against original video recording - multiple analysis cycles of empirical data adds to prolonged engagement

Rich thick description (Creswell, 2003)

- plausible and context rich account - use of participants quotes to illustrate concepts

Member check (Dallimore, 2000)

- following the interviews, participants were provided with a transcript of the interview

Persistent observation(Lincoln and Guba, 1985)

- site visits wherever possible

natural history test of validity (Neuman, 2000)

- philosophical assumption and method clearly articulated in a logical manner

Triangulation of evidence (Patton, 1990; Creswell, 2003)

- converging conclusion from triangulation of method, data and theory

Triangulation 4.4.1

Presenting information that counters themes(Creswell, 2003)

- theoretical scaffold encouraged seeing data from multiple levels of abstraction.,- Critical Realism and agency theories present congruent but different perspectives

Dependability (Reliability in Positivism)

Reliability audit, interview protocol, establishment of a case study database (Eisenhardt, 1989; Merriam, 1998).

- question-bank - interview protocol -case study database

Protocols 4.4.3

Clarifying researcher bias(Merriam, 1998; Creswell, 2003)

- making clear the researchers’ assumptions about her world view and theoretical orientation

Confirmability audit (Lincoln and Guba, 1985)

- record of data collected (including videotaped interview and transcripts) as a part of the chain of evidence. - a commitment to presenting transcript material with as little editing as possible

Chain of evidence 4.4.3

reflexivity (Lincoln & Guba, 1985)

- extensive use of memoing to monitor any personal bias, research changes

Reflexivity 4.4.1

Use of established procedures (Miles & Huberman, 1994).

- articulating procedures

Transferability (Ext Validity, generalisability in Positivism)

Thick description (Corden and Sainsbury, 2006)

- findings use verbatim quotes extensively, to illustrate the actual data emerging from the study

adequate amount of data (Morse, 1994)

-multiple sources of evidence -consideration of sampling strategy

Use of established theories; Tsoukas 1989; Yin 1994)

-MTS is used to understand interactions between the environment and the agent

Transferability of findings -conceptual frameworks applied to other domains (for example, mobile phone and education; see 7.5)

Confirmability (Objectivity in Positivism)

validity of subjective interpretation (Miles and Huberman , 1994)

High level of reflexivity, bracketing, tools such as researchers diaries. Citing research on bias to show awareness.

Reflexivity 4.4.1

audit trail (Miles and Huberman, 1994) or "inquiry audit" (Lincoln and Guba, 1985, p. 317)

Audit trail clearly shows the framing of the study and collection of data, Development of diagrams demonstrates the internal consistency of the information collected. Nvivo file captures progressive coding.

rival explanations and alternate instances (Miles and Huberman, 1994)

Use of multiple theories in MTS, abstracted mechanisms grounded in empirical data.

Explanation 4.4.1

APPENDIX 7: CASE STUDY ORGANISATION VIGNETTES

* Name not provided to ensure participant anonymity ** large multinational Sample Organisation ID Description Addendum

1 Krome B Krome Studios http://www.kromestudios.com

Primary product and service: One of Australia’s largest video game developers and a leading independent development studio worldwide. Products include TY the Tasmanian Tiger™ (Krome’s own IP) Employs: 300 full-time employees spread across studios in Brisbane, Melbourne, and Adelaide. (see addendum).

Significant number of staff ‘let go’ during 2009 and 2010. See http://www.tsumea.com/australasia/australia/news/151010/serious-trouble-at-krome-studios

2 THQ Australia H THQ Studio Australia http://www.thqstudioaustralia.com International games company with local studios.

Closures Aug 2011 http://www.smartcompany.com.au/buy-or-sell-a-business/20110811-video-game-industry-hit-as-thq-closes-two-australian-studios.html

3 Creative Assembly

G The Creative Assembly http://www.creative-assembly.com.au

Primary product and or service: Game maker specialising in strategy war games.

Sega buys company in June 2011

4 Pandemic F Pandemic Studios http://www.pandemicstudios.com

Primary product and or service: creators of Destroy All Humans!™,

Brisbane studio closed late 2009.

5 * K** 1

Multimedia

6 Ice Media A Ice media http://www.icemedia.com.au

Primary product and or service: Icemedia is the largest website, multimedia and applications design and development company in Queensland and one of the largest in Australia. Employs : Marketing consultants, graphic designers, information architects, usability consultants, interaction designers, software engineers, copy writers and editors, web developers

7 Queensland Transport

8 Queensland Museum

9 * C 1

10 ABC Online News

I ABC Online- Brisbane http://www.abc.net.au/news Primary product and service: ABC Online has more than 1.7 million pages of content across 15 subject gateways, with additional services offered via WAP, SMS and i-mode mobile formats.

11 BCM J BCM http://bcm.com.au, blogs.bcm.com.au , twitter.com/bcmpartnership Primary products and services: producing interactive media and other media formats Employs: 80 Specialist staff, based in both Sydney and Brisbane offices,

APPENDIX 8: CASE STUDY AND INTERVIEW PROTOCOLS

CASE STUDY PROTOCOL

The case study protocol contained the following items.

o An overview of the project (project objectives and case study issues).

o Field procedures (credentials and access to sites).

o Questions (specific questions that the investigator must keep in mind during data

collection).

o Guide for the report (outline, format).

o Management of data and confidentiality.

o Digital encoding requirements.

o Examples of secondary sources of evidence.

INTERVIEW PROTOCOL

1. Prior interview

a) A letter was sent via email to each participant prior to their scheduled interview. The letter

provided an overview of the project and the consent form. Areas of concern such as permission to

enter the site were mentioned, where agreement to the interview indicated approval for the

researcher to enter the workplace. This letter clearly indicated the researcher’s credentials. Attached

to the letter were the following.

b) A link to a website with information for participants

http://www.anitzageneve.com/qut/project.html.

c) An attachment of the Australian Culture and Leisure Classifications (ACLC) ‘Interactive Content

Creation’ class 267 (Australian Bureau of Statistics, 2008) definition. This is provided to potential

interviewees so that they can confirm their suitability. Further clarification of eligibility occurs

through a discussion with the interviewer.

d) Flyer (which could forwarded to other possible participants for example, snowball sampling)

Item b) (below)

Item c) Extract outlining the Interactive Content Creator occupational role

The interactive content creation class (267) consists of units mainly engaged in the

development, creation and production of interactive content for digital products and services.

Interactive products are those for which the narrative is able to be controlled or influenced by

consumer-participants.

Exclusions / References Units mainly engaged in;

(a) producing animated films are included in Class 263 Film and Video Production; and (b)

developing commercial business application software, such as word processing and spreadsheet

applications, are out of scope of the classification.

Primary Activities

Interactive content design and development

Interactive film production or post-production

Interactive games design and development service

Production or post-production of multipath films and other interactive style films

Web site design and development service

Source: ABS (2008, pp. 49)

Item d) (below)

2. Interview protocols:

Interview stage Requirement

Prior to interview Email an overview of the research project, their requirements as participants and the pre-interview questionnaire (see Appendix 8 for online questionnaire)

Advise employer if required of onsite visitation, include overview and credentials Email participants a reminder for the time, date and location

At interview Confirmation that informants occupational role is suitable for the sampling strategy

Interviewee advised of time schedule- 45mins Explanation of confidential and non identifying information Awareness of reflexivity - interviewee expresses what interviewer wants to hear Reiteration of purpose, researcher’s credentials, timing Style of interview is personal and unbiased Self-disclosure -should be avoided if it seems like it's leading to interviewer bias Broad initial question: I’m interested in knowing your thoughts on what influences there might be on women’s participation in the Digital Content Industry. If the participant did not respond or required clarification this question was reposed as. Tell me about your experiences in the industry, how did you get here? What have been the influences? A list of probes in instrument General probes: probes non-leading but not necessarily dispassionate. For example, “Really (to show interest), can you tell me more about

that? Why do you think that was (in reference to described scenario)”. Specific ‘participant’ prompts: Used only when the interviewee raised a related point. For example, “I’m not sure I understand, did you mean x, y, z (very accurate rephrase) with link to research aspect”. Relational prompts: As the researcher could readily be identified ‘as one of the participants for example, women working in industry, reference could be made to the researchers own experiences. Likewise, researcher had a child so could draw on that experience to connect with participants. Focusing/investigative prompts: when a researcher felt that something has been touched on or hidden in the participant’s prior comments. For example; “I’d like to come back to something you mentioned earlier...”(see question bank)

Following interview Digitise footage as soon after the interview as possible.

Ensure that the transcript does not identify the participant Store interview transcript in a secure and safe manner. Following the interviews participants will be provided with a transcript of the interview via email.

Field notes/ observations

These were recorded on paper and entered into the case study database (Nvivo project) as soon after the observation, where possible observations were memo’s linked to particular participant interviews.

3. Management of data and confidentiality

Pseudonyms were used for data arising from questionnaires and interviews

Questionnaire and interview transcripts were not tidied up, as recommended by

Corden and Sainsbury (2006); consequently spelling errors in the questionnaire and

use of colloquialisms remain intact.

Digital encoding requirements. Interview data was recorded on video, which

potentially has a higher risk of the participant being recognised. Video was viewed

only by the researcher (that is excluding supervisors). Recordings in Phase 2 were

outsourced for transcriptions. In this instance all video was converted to audio before

being submitted to the transcription provider. Naming of files to be transcribed by

third party service

participant number_industry sector_phase

for example, 2. games_phase 1 is the second participant from the games industry

o In regards to storage and disposal, recorded video was encoded to DVD and

password protected and these were kept in a locked storage cabinet only

accessible to the researcher.

4. Ethics.

As this research involved the active participation of human subjects for example,

questionnaires and interviews, it required ethical approval to meet standards of the

Queensland University of Technology policy and Commonwealth and State

regulations / legislation. In accordance with institute policy, documents included a

participant information sheet, written consent from participant and ensuring

anonymity of participant.

APPENDIX 9: ONLINE QUESTIONNAIRE

The following questionnaire was administered using survey monkey. Completed

questionnaires were sent to the researchers email account.

Name: Age:

Gender:

Current role and organisation:

Your average workload per week is:

Length of time in role:

Length of time at current organisation:

Previous work: type of organisation and description of job:

Length of time in Digital Content Industry:

Number of employees in your current organisation: Males: Females:

What are you qualifications and the year completed:

How would you describe your current position in terms of quality of experience (for

example, fun, exciting, fulfilling, repetitive, and boring)?

Describe the types of activities you typically do in your job? for example, program, design,

manage people, data entry

Do you believe there are factors that are barriers for women to participate in the Digital

Content Industries? Answer values range from 1 (strongly disagree) to 6 (strongly agree):

Working in your current position or similar what factors have you personally experienced

that have been a negative influence on your career/participation? Please list or describe the

factors:

Based on previous experience within your industry what percentage of males and females do

you think occupy the following roles?:

a)Junior programmer b)Senior programmer c)3D modeller d)HTML

coder e)Graphic designer f)Other (project managers) g)Other (please type)

The following question is about your exposure to certain media and resources. Where did/do

you source your information about employment prospects?

a)Online b)Recruitment agents c)Verbal- peers, network d)Work

Where do you source your information about the industry? Examples include:

a) Magazines- industry specific b)Newspaper c)Online web sites d) Blogs

e) Industry events f) Training events g) Exhibitions

APPENDIX 10: EXAMPLES OF INTERVIEW QUESTION BANK

Part A- Past Question Prompts/Possible responses Literature Can you tell me the story of how you

started in the industry?

Prompts: Did your experience of computers at school influence your occupational choice? Did someone you know provide an interest or opportunity? For example, spouse. Possible Answers: role models for example, parents, partner

Wisenet and CQU survey asks these questions and nearly a third of the women interviewed went into industry not because they were recruited into it, but because they felt unwelcome in academia (CEN Report, 2001http://pubs.acs.org/cen/employment/7927/7927employment.html)

Do you recall whether you thought there were going to be barriers when you started in the DCI?

Did you have a clear idea of what you wanted to do when you entered the industry?

Why did you enter the industry?

For example, Financial, Travel, Flexible work hours, Challenge- changing technology

Pantelli (1999) asks this question in UK and finds ‘want a challenge’

What was your previous role/organization and what were your reasons for leaving your previous employer?

Part B- Present

What do you enjoy about your current role?

How confident are you in your role? McGrath-Cohoon and Aspray (2006) suggests confidence may influence women to leave the IT profession

In your pre-interview response you mentioned you had experienced certain factors such as … …tell me more about how you responded to these.

If you have experienced x negative factor why do you continue to participate?

Have you ever considered leaving the industry due to concerns about employment options? Why?

Pay inequity prompt, contract versus fulltime

Why do you stay in the industry?

For example, find it interesting- what aspects. Don’t know what else to do? Trying to leave- why?

Part C Media Where did/do you source your

information about employment prospects?

For example, Newspaper Online Verbal- peers, family, university, work

Where do you source your information about the industry?

For example, Magazines industry specific Magazines women’s interests, Newspaper, Online, Blogs, Industry events, Training events

Such documents may be used as multiple sources of evidence for case study (Yin, 1995).

Part D: Revisit research questions

In your response you said you had experienced an event that you would consider a negative influence on your career? Can you tell me more about this event.

The event was__________________

Do factors exist for women in the industry? Please indicate if you have experienced these personally.

Do you think there are specific factors that can influence women in ICT-DCI compared to other industries?

Women see themselves as different to other women (Cukier,Shortt and Devine, 2002, Betz and Hackett (1997)

Part C- Future

What long term goals do you have? For example, personal goals (choice content and performance)

Do you think you will face barriers in the future?

SCCT (Lent et al., 1994) -outcome expectations

How committed are you in staying in the ICT industry?

How would you describe your ideal position?

Do you think you have the appropriate skills to achieve this?

Do you know of any support options that are available to encourage women to participate in ICT related industries?

For example, WIT Office for Women

ST (Giddens, 1984) suggests allocative resources may be an influence on agency.

APPENDIX 11: HERMENEUTIC PRINCIPLES APPLIED IN THE RESEARCH

How the research process addresses the principles proposed by Klein and Myers (1999). Principles Its methodological emphasis Implemented in the current study

1. Hermeneutic tradition

Explain the nature of socially constructed human meanings and the interdependent meaning of the parts and the whole that they form

This principle is achieved by iterating between considering the shared meaning of the parts (influences, interactions) to the whole (the participation). The aim of understanding the deeper aspects of participants accounts/

2. Contextualisation

There a need to explain the socio-historical context and how this context has evolved to its current state

The context of the study identifies; - the DCI, in specific the multimedia and games industry in Queensland - the historical context of gender and social equity issues

3. Interaction between researcher & subjects

Explain the mutual interactions of the researchers with the participants.

A high level of reflexivity is a goal of the research and this includes a clear articulation of the researchers experience in the industry sectors and personal motivations. A reflexivity log/diary is maintained.

4. Abstraction & generalisation

Explain how insights are derived through the use of a particular theoretical perspective acting as a sensitising device to view the world in a certain way

Clear description of the perspective to be applied to the research problem including the Critical Realist philosophy, and also the theoretical scaffold comprised of agency theories.

5. Dialogical reasoning

Explain possible contradictions between the theoretical preconceptions and the data gathered, allowing for a subsequent cycle of revision

The use of multiple theories as an intellectual basis of the research is made explicit.

6. Multiple interpretations

Explain possible variations in participant’s interpretations.

The Individual Differences Theory expects there to be variation in narratives of experience. However, the goal is to also look for any key commonalities. Collecting a range of data from various sources and the use of multiple theories for interpretation also lends itself to multiple interpretations. The use of several theories, which offer different perspectives, also encourages a range of possible explanations.

7. Suspicion Explain possible bias in narratives collected from the participants and in finding alternative explanations to the problem under investigation.

Examine the views and actions of participants and not of the researcher. Use of rival and multiple theories. High level of researcher reflexivity.

APPENDIX 12: GIDDENS’ MODEL OF INTERACTION (BASED ON GIDDENS, 1984)

Structure Signification Domination Legitimation

Modality Interpretive scheme Facility Norm

Interaction Communication Power Sanction

Key objects signification or meaning making (interpretive schemes used in the communication of meaning)

power and control (facilities used in the exercise of power relations)

regulation and sanction (norms applied to the sanction of social behaviour)

Description Agents draw on structures of "signification" (language and other symbolic codes) to produce/reproduce "communication" (meaningful and understandable interactions) via the modality of "interpretive scheme" (stocks of knowledge).

Persons exercise "power" by accessing "domination" structures, for example, allocative (control of materials like money) or authoritative (control of people) through the modality of "facility". Power is not fixed but a capability.

Agents "sanction" each other by drawing on "legitimation" structures (moral beliefs) via the "norm" modality

APPENDIX 13: SEARCH STRATEGY DETAILS

The choice of studies to include in the early literature review was driven by the following

three key reasons.

The study provides a general insight into the DCI and ICT context,

The study identifies influences on participation explicitly relevant to women

The study include an empirical element (where the cohort is women working in the ICT or

DCI industry) or a conceptual element (offering a theoretical perspective).

The articles in the review were initially sourced from top ranking journals as identified in Table 61. For

example, Ulrich’s indicates if a journal is peer reviewed and QUT library resources

(http://www.library.qut.edu.au/subjectpath/citation_indexes.jsp) provides general links to citation

indexes. Journal articles were also referenced from Web of Science, which uses a citation index

(Institute for Scientific Information ISI) citation. This provides some guarantee of the quality of papers,

compared to general web-based search engines. Similarly, the Journal Citation Reports (JCR) provides

an indication of the ‘impact factor’ of journals; however, this was not available for the entire range of

publications. Consulting the references list of these articles allowed for a snowball approach to identify

further articles. Searches within key databases included keyword searches for “Gender, Gender and

ICT, Gender and IT, new media” and Boolean searches combining these.

Table 59 Example of key databases consulted for literature review Type Specific reference

Online journal database IEEE Xplore Proquest(ABI/INFORM)

Web of Science

ABI/Inform

ScienceDirect

ACM (Digital Lib) JSTOR

Government reports NSF (USA), DCITA (Australia)

Each document used in the review was stored within an NVivo project as either a full text

version of an article or a proxy document. A casebook was created to enable the following attributes to

be coded for each document (Table 62). Each case was coded into the nodes identified in the table.

Table 60 Research design attributes identified in the meta analysis of key literature.

Node title Attributes Source Example include journal, conference, book

Research type Conceptual, empirical

Epistemology Interpretive, Positivist, Critical

Methodology Case study, Interview, Survey

Type Primarily explanatory, exploratory or descriptive, conceptual or empirical

Country of study USA, UK

Context Domain the research purports to, industry field

Cohort If empirical study-cohort sampled

Source Name of Journal

Source type Journal, conference, book or other

Author

Influences

APPENDIX 14: EXAMPLE OF RESEARCHER REFLECTION

A reflexive approach can be applied to interviews (Silverman, 2009). Although the

interviewer aimed for a consistent interview style, upon reflection it was noted that interviews had a

different feel to them. Sometimes this was due to the personality of the interviewee, others it had to do

with the environment for example, cold, rainy day. There was a concern by the researcher that these

factors may influence the types of questions being asked. Upon further reflection, the researcher can

see the different types of questions asked were consistent and appropriate. As Table 63 indicates, the

questions in the interview included the different types recommended by Strauss, Schatzmann, Bucher,

Ehrlich and Sabshin, (1964).

Table 61 Types of interviews questions (based on Strauss et al., 1964, pp, 26 – 27) Type of question (based on Strauss)

Example in the reported research

The challenge or devil’s advocate question

A participant stated there were no barriers to women’s participation. The researcher then asked, “why then are there only two women out 50 employed in your organization?".

The hypothetical question The researcher asked “If you had to give advice to a young girl what would you describe as the positive aspects and the negative aspects of working in the industry”

Posing the ideal The researcher asked, “What could we do to encourage more women into the industry?”

Offering interpretations or testing propositions

The researcher asked, “So are you saying that you lacked confidence?”

Researcher’s memos -Following are two examples of the researcher’s memos.

June 2007 I’ve just finished coding the first 12 interviews in NVivo. There are over 150

nodes or possible ‘influences’! The only theme that stands out is that of lifespan. I'm now going to sort

them into the E the P and the dash categories. The environment seems easy, as it's mostly descriptive of

structures such as settings, the P seems harder but it is seems to involve labels surrounding identity.

October 2008 I realised today there was a tension between the gendered identity and the

occupational one. I had been so focused on seeing their experiences as ‘women’ that it took a while to

'hear' the data. I reflected that there had been a distinct pulling away from a gendered identity when I

first contacted them/meet them. I had to sell the study to them as a study of 'industry professionals'

rather than women. I have been describing the individual as a gendered but that’s the problem! When

we see a female DCI worker we see the female and not the DCI.

March 2009 – spent some time thinking about how I collected the data.

Interviewer as native- It appeared to the researcher that as participants perceived her to be

one of them (a member of the DCI community and also a female) that this facilitated access to

participants and an open sharing of personal information during the interviews. However, one early

interview highlighted to the researcher the need for neutrality in positioning the research to

participants. During the interview, the researcher noted an initial standoffishness by the participant,

indicated in most by the participant’s negative body language and coyness in responding. This was in

contrast to previous interviews, where the researcher had readily established a repoire with

interviewees. Over the course of the interview, the interviewee appeared to relax, even commenting at

the end how enjoyable it had been. When the researcher enquired for the reason of the initial

awkwardness, the interviewee replied in a somewhat relieved fashion that she had thought the

researcher was a “feminist”. This led to the researcher’s reflection on how participants perceive gender

studies.

Secondary data sources- Secondary sources were chosen because they were mentioned by

participants. For example, a participant mentioned a website Gamasutra and this site was then treated

as data and analysed. The exploratory approach to secondary sources of data led to several ‘unplanned’

sources of data. One interesting scenario arose when the researcher attended the Game On: Wired

Women event (Item 49) and expected that an audio transcript would be made available of this event;

however, this did not occur. Few field notes had been recorded in light of the planned recording.

Following up this event in the online environment led to the identification of a website blog discussing

the event. This blog was then treated as a secondary data source. Another example emerges from the

comments participants made towards the gendered nature of games products. This then led to the

researcher analysing the types of games products available. Other key secondary sources of data

included over 1000 raw comments from an international survey of game professionals (see IGDA,

2005a, b), several were from Australian respondents. Hence, as Layder (1989) recommends there was

certain flexibility towards the collection of multiple sources of data. This was seen as a fruitful and

beneficial approach.

APPENDIX 15: METHODS OF DATA REPRESENTATION USED IN THE CURRENT STUDY

Table 62 Approaches to data representation implemented in the study Representation method As implemented in the study Graphic (see Larkin and Simon, 1987; Bauer, 1993) and Integrative diagrams (Strauss, 1987) Visual model (Creswell and Brown, 1992)

Findings conceptualised in several different formats including matrixs, tables and illustrations.

Conceptual Framework (Miles and Huberman, 1994, pp. 18-22)

Models such as the Sphere of Influence organise data.

Matrix (Miles and Huberman, 1994) Early on in the data process, a matrix of influences encountered by participants was developed. Ordered matrix’s based on data categories were more useful than unordered matrix’s.

The ‘Acts of Agency’ are presented in a tabular form with participant’s comments to illustrate these constructs.

Conditional matrix (Strauss and Corbin 1990, pp. 255-256).

Matrix listing the environmental characteristics or conditions (see Table 36).

Time displays (Miles & Huberman, 1994, p. 91)

Lifespan concept within the Sphere of Influence (see Table 42). Used as a descriptive organising device for rendering a sequence of events more comprehensible.

Typology (Patton, 2002)

The MTS typology of theories involves the categorisation of theories based on their level of abstraction.

b. Danermark et al.’s six stage model Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6

Following are several examples of early data representation that occurred during research

Existing research informing the process a) Layder

Adaptive Theory offers an extensive argument for using theory as a scaffold, congruent with Critical Realism

b) Danermark et al. Offer a specific methodological approach for data analysis, congruent with a Critical Realism

c) Bhaskar As the key theorist for Critical Realism offers ‘three domains of reality’

Models emerging from the research process 1. SoI- Sphere of Influence

Sensitises researcher to characteristics of the environment (E) and person(P) EoI - Events of Interaction

Sensitises researcher to specific events of interaction (between E and P) MTS- Multi-Theory Scaffold

Guides choice of existing agency theories used to stimulate researcher’s insights into underlying mechanisms

Five Acts of Agency Identifies agent-driven mechanisms involved in the interaction between E and P

Acts of Agency theory

Example of unordered matrix used during data analysis Influence Phase 1 Participant ID

m1 m2 3 4 5 6 7 8 9 10 11 12

Social agents x x x x x x x

Access to resources x x x x

Role models x x x x x

1. SoI 2. EoI

3. MTS of Agency Theory

Mechanisms

c. Bhaskars three domains of reality Empirical Actual Real

Framework 1-analytical (inductive)

4. Five Acts of Agency

a. Layder’s Adaptive Theory (scaffold concept)

5. Acts of Agency theory

Framework 2-theoretical (abductive)

Framework 3- ontological (Multiple logic including retroductive)

APPENDIX 16: REFLECTION ON THE APPLICATION OF GIDDENS’ STRUCTURATION THEORY (ST)

Table 63 Application of Giddens’ ST (Giddens, 1984, pp.281-28) and Shoib et al.’s (2006) recommendations towards applying Giddens ST

Giddens’ recommendation As implemented in this investigation 1 All human beings are knowledgeable agents Informs epistemological approach, suggesting

the appropriateness of hermeneutic analysis

2 The knowledgeability of human agents is always bounded on the one hand by the unconscious and on the other by the unacknowledged

Recognises the limitations of participants subjective insights

3 The study of day-to-day life is integral to the analysis of the reproduction of institutionalised practices

Informs epistemological approach with a focus on everyday lived experience

4 Routine, psychologically linked to the minimising of unconscious sources of anxiety, is the predominant form of day to-day social activity

Similar to Bandura and Beck with anxiety and risk minimization strategies by the agent

5 The study of context, or of the contextualisation of interaction, is inherent in the investigation of social reproduction

Case study identifies a context and the unit of analysis is the everyday social practices inherent in the interaction between agents and their environment

6 Social identities, and the position-practice relations associated with them, are ‘markers’ in the virtual time-space of structure

This highlights the importance of identity as a marker.

7 No unitary meaning can be given to constraint in social analysis

The constraint of an influence is in the eye of the beholder

8 Among the properties of social systems, structural properties are particularly important, since they specify overall types of society

Sensitizes the researcher to this level of analysis

9 The study of power cannot be regarded as a second-order consideration in the social sciences

This sensitizes the researcher to the concept of power as a possible underlying mechanism

10 There is no mechanism of social organisation or social reproduction identified by social analysis which lay actors cannot also get to know about and actively incorporate into what they do

Participants may have some knowledge of the underlying mechanism

Shoib’s recommendation As implemented in this investigation 1 Sensitivity towards the context properties of systems Clear definition of the study’s context

2 Time insights into the immediate environment and past experience of the social actors; and the broader social and longer-term historical setting.

Acknowledge history of both longer duree of society and individual histories

3 Examples of the ongoing production and reproduction of practices through social routines and un-stated cultural norms

Level of analysis may consider social practices.

4 Aspiration to overcome the division between macro and micro analysis

Level of analysis must consider both macro and micro aspects.

5 Complexity of human intentionality, deliberate actions Agents are not passive, therefore they must do something, and they must act.

6 Subtlety of structural constraint May consider subtle or hidden influences.

APPENDIX 17: SECONDARY SOURCES OF DATA

Following is a list and description of all the secondary sources data used during data analysis and further theorisation. The last column, ‘PDF’ indicates if a PDF of the data is included in this dissertation.

No. Title Description Date Industry Key Purpose for inclusion PDFattached

Item 1

Inquiry into pay equity and associated issues related to increasing female participation in the workforce

162 submissions to the Australian House Standing Committee on Employment and Workplace Relations, Committee

2008 Wide range A wide range of other industries, where women’s pay equity has been raised as a concern. There is a notable lack of submissions from the Australian DCI. http://www.aph.gov.au/house/committee/ewr/payequity/

Item 2

International Games Development Association (IGDA) website

Greater Brisbane Chapter of International organisation website

2009 Games Provides background to international games industry context. Highlights types of industry events being promoted to those people interested in the games industry, including the 2009 'GameOn' series of public lectures. Links to other websites of interest. Indication of organisations in Australia. http://www.igda.org/

Item 3

Game Developers' Association of Australia (GDAA) website

Australian based games developer organisation

2009 Games Provides background to Australian games industry context. http://www.gdaa.com.au/

http://www.gdaa.com.au/careers/women.html

Item 4

Australian Interactive Media Industry Association (AIMIA)

State Chapter presidents 2009

2009 Multimedia Provides background to Australian multimedia industry context. Australian Interactive Media Industry Association (AIMIA) website (screenshot 16th Oct 2009).

All four Chapter presidents are male. Male majority on the QLD

committee. Not a measure but an insight how the industry image remains “masculine” simply due to the ratio. http://www.aimia.com.au/i-cms?page=1

Item 5

Games company on Facebook

Halfbrick Facebook page

2009 Games

http://www.facebook.com/halfbrick Screenshot 18th Oct 2009

Item 6

GDAA website/ Government sites advising on careers

Special Section on careers and women sponsored by Knowledge Skills Victoria

2009 Games Front page identifies content as including “discussions about women in the industry”. Identifies the presenters of the 2004 GDAA conference. Male majority, of the female’s one is a politician and another, the Chair. Describes the IGDA website as a "well resourced site"., however links suggest otherwise where the link to a "girls guide to gaming” is essentially a PDF of outdated links (for example, to 2004 stories/news) and other websites (see item 9) http://www.gdaa.com.au/careers/index.html http://www.gdaa.com.au/careers/women.html

Item 7

E3 website Images of Booth babes 3/2009 Games Sexualised images of women as booth babes or entertainment at games events such as E3.

http://www.e3girls.com/

Item 8

Girls Guide to Gaming A PDF document with links to websites with a short review of the site.

3/2009 Games Makes reference to the site not being "girly". States that it aims to dispel myths that the industry "is run by geeks". Provides resources for "children". http://www.gdaa.com.au/teachers-forum/Documents/Girls_Guide_To_Gaming.pdf

Item 9

IGDA Women in Games SIG website

Online article 3/2009 Games Headline articles such as "why it is critical more women than ever play and develop games', "why we need more women". Well emphasised links to social networking sites such as LinkedIn, for "occasional job postings". http://www.igda.org/women/resources.html

Item 10

EDGE online magazine (by line: The Global Games Industry Network)

Article on Jade Raymon (Ubisoft) Article "why we need more women in games"

3/2009 Games Article reports on a female role model in the games industry. Comments responding to the article range from disparaging to supportive. Reports on a GDC panel who suggested that "women don't always recognise or even have confidence in the ability to be games designers'. http://www.next-gen.biz/news/jade-raymond-to-manage-ubisoft-toronto

Item 11

GameDev.net website Article title "Play games to work smarter: why it is more critical than ever that women and play and develop games"

4/2009 Games Reporting on the IGDA Women in Games SIG refers to cyber-socialisation, serious games in areas were women are dominate for example, healthcare, "obstacles exists" when "young women come in, they often sit in the back and think that they don't belong" These young women cannot "imagine themselves as games designers". http://www.gamedev.net/columns/events/gdc2009/article.asp?id=1764

Item 12

The Glasshammer website

Article "what am I worth" 3/2009 ICT Personal account of a woman receiving 'constraining' feedback and an agentic response. http://www.theglasshammer.com/news/2008/10/22/what-am-i-worth/

Item 13

EDGE online magazine (by line: The

Article title "Microsoft recognises games

3/2009 Games Reports on the "first ever Women in Gaming Awards" luncheon, as being "an intimate event".

Global Games Industry Network)

industry women" Article title “GDC: We need more women in games”

http://www.edge-online.com/news/microsoft-recognizes-game-industry-women

Item 14

Promotional material for games company

Pandemic promotional leaflet distributed at AGDA 2006 conference

2008 Games Describes the studio environment as being an eclectic mix of colleagues and friends who share interest such as partying, surfing, camping to gaming and poker.

Item 15

Pink Collar report Government whitepaper 2007 Technology Describes one initiative in the education sector, the "Diploma of Multimedia for Women" program. This same program in the early years charged the female participants an extra $500 (in comparison to the non women’s Diploma) for the guest 'role model'. One interview participant went through this program and described her experiences.

N-difficulty in finding a copy, contacted author Sonja Bernhardt

Item 16

Screen Goddess calendar

Calendar and blogs commenting on calendar

2005 ICT A calendar where the sexualised images of female IT professionals caused controversy. http:// http://www.itgoddess.info/ Further resource includes a blog responding to the calendar, which notes the withdrawal of funding from the Australian Computer Society. http://rosemary.id.au/view/blog/hullabaloo/ Blogs: MashUp / Cheesecake calendars Archives Helen July 12, 2006 05:24 PM The Sydney Morning Herald http://blogs.smh.com.au/mashup/archives/your_say/005178.html

Item 17

Women in IT UK email newsletters

Email from industry group for women

2008-2009

ICT Email asks "do you lack confidence". Y

Item 18

Women in Games in Australia website

A website promoting women in games.

2008-2009

Games Little activity is evident on this website. http://www.womeningames.org.au/educators.htm

Item 19

Desktop magazine Australian Design:Australian Culture magazine issue 250 June 09

2009 Multimedia Competition features an illustration of a semi-clad woman on fire. Competition in supported by several education and industry groups for example, QaNTM, Wacom, Semi-permanent conference.

Item 20

Google search Search term “warez download dreamweaver”

2009 Multimedia Googling a search term that a typical user looking for free software may use. The fifth search return led to a website that offers several links in the navigation bar to software, ebooks and then live sex. One website has a markedly feminine look to it, which contrasts the majority that are seemingly targeted towards men.

The fifth search led to a website that offers several links in the navigation bar to software, ebooks and then xxx and live sex

Item 21

Job advertisements LinkedIn job advertisement

2008 Multimedia/Games

Several job advertisements feature the word “passionate" as a characteristic of the DCI worker.

Item 22

Blog Ashley Jenkins pro-gamer

3/2009 Games Female games pro-gamer personal blog makes the statement that she looks “forward to when gender is no longer an issue”. http://www.news.com.au/ link no longer available

Item 23

Website featuring female role model in the games industry

Website 2009 Games Gamasutra website and Edge online magazine rank the top role models in the games industry.http://www.edge-online.com http://www.edge-online.com/features/game-industrys-100-most-influential-women?pa

Item 24

Google search Search term ‘jobs in the games industry’

16/8/2009

Games Google search 'jobs in the games industry' limited to Australia search data. Search returns a predominance of design positions. Links to jobs advertisements specifying passionate people. August 16th 2009 2nd Google item linked to the following website http://www.jobs.com.au/gaming-industry-jobs/ Suggest industry requires passionate and creative individuals.

6th Google search result http://www.simplyhired.com.au/job-id/3omsdz3xj4/art-director-jobs/

Position offers mentoring and collaboration. Advertisement states the position will provide 'mentorship, coaching and performance management of lead artists on a game team as well as team members in collaboration with the Lead artists"

Item 25

WomenGamers.Com - Article in Online Gaming Magazine ‘Gaming Magazine Ads: Failing the Female Market’.

3/2010 Games Blogger discusses the low levels of women readers of games magazines and attributes this to the images (advertising) of women in the magazines. http://www.womengamers.com/articles/editorials/gaming-magazine-ads-failing-female-market/ Review by womengamers.com on games magazine ads. Highlights the stereotypical and sexist portrayal of women and girls in the advertisements.

Item 26

WITI (organisation aiming to support women in IT careers)

Email promoting training opportunities

9/2009 ICT Training targeted towards women in the IT industries. Highlights key words of ‘passion’.

Item 27

Eweek magazine Article in online magazine

2008 IT Online article discussing wage gap for women in IT http://www.eweek.com/c/a/IT-Management/Wage-Gap-Narrower-for-Women-in-IT/

Item 28

Diversity Report 2006 Games Provides 1000+ unedited comments from the survey. http://www.igda.org/diversity/IGDA_Comments-on-Diversity_Jul05.pdf

Item 29

Gamasutra website Times series of front page

2009 (archive 2001-2008)

Games Historical time series of a key games industry website. Time series involved a screen capture in the first or second month of each year from 2002-2008. In 2002, the webpage includes an image labelled ‘negotiating contracts'. It is a retro black and white photograph featuring only men at the negotiating table. Distinct lack of female presence in authors, contributors, reviewers. Graphics targeting men. Job ads seeking 'skilled and passionate' people. By 2008 the side bar features older women at a games event. http://web.archive.org/web/20010108104100/http://www.gamasutra.com/

Y-see next table

Item 30

ComputerArts Magazine (Print) 2010 Multimedia Key industry magazine. UK based but has an Australian publication. Several articles illustrate a male ratio for example, of the featured creative’s one of nine is a women, the article ‘how do you make a good brand great’ has six authors(all male), six judges of a design competition (all male).

Item 31

Blog Technology Blog where several women are discussing why there aren’t more women in technical roles. Points out that there are no ‘barriers’ but rather women feel left out. http://www.lealea.net/blog/comments/women-in-tech-asking-the-wrong-questions/#ixzz0cRlJSESg

Item 32

Article Women ‘not real techies’ 3/2010 ICT http://www.womenintechnology.co.uk/news/stop-the-women-techie-snobbery--news-19566445 The UK based ‘women in technology’ website reports that women in non technical roles in IT are being described as not “being real women” in IT.

Item 33

Computerworld online News website 3/2010 Games An online ‘general media’ article, written by academic Mary Pratt, and titled ‘Computer game industry looks to women for fresh insights” http://www.computerworld.com/s/article/293317/Computer_game_industry_looks_to_women_for_fresh_insights?taxonomyId=10&intsrc=hm_topic

Item 34

GamesSpot Online article 3/2010 Games EA spouse case http://au.gamespot.com/news/2004/11/11/news_6112998.html EA:The human story. http://ea-spouse.livejournal.com

Item 35

Women in Technology UK Email

Email via LinkedIn group promoting training opportunities to women in IT

6/2010 ICT Email announcing an evening training course for women. Highlights the need for negotiation and “Why it's necessary for women to negotiate successfully”. This includes negotiating salary. The subject line reads, “why women don’t ask”.

N - private email

Item 36

AWE and ETHE are surveys of employers while EEBTUM is a survey of employees.

ABS surveys calculate gender pay gap.

3/2010 Range The quarterly Average Weekly Earnings (AWE)104106, the annual Employee Earnings, Benefits and Trade Union Membership (EEBTUM)105F107 and the biennial Employee Earnings and Hours (ETHE)10108 surveys suggest women earn less then men.

Item 37

Online magazine article ‘Builder AU’

Games development: a real career choice?

3/2010 Games Website discusses working conditions in Australian games organisations. Spotlight on poor working conditions.

106 ABS, Average Weekly Earnings (Cat. No. 6302.0), May 2008. 107 ABS, Employee Earnings and Hours (Cat. No. 6306.0), May 2006 108 ABS, Employee Earnings, Benefits and Trade Union Membership (Cat. No. 6310.0), August 2007

http://www.builderau.com.au/strategy/futuretech/soa/Games-development-a-real-career-choice-/0,339028285,339169404,00.htm

Item 38

Website advertising remote workers

3/2010 Games/ Multimedia/ICT

Advertising that Australian companies could replace their Australian workers with remote workers and only pay $4-$10 dollars an hour for remote workers. This indicates the global pressures on local workforces.

http://www.remotestaff.com.au/ Item 39

‘Good game’ website to accompany ABC broadcast television channel

Website and archived video of television show on gaming

2006-2009

Games First aired 2006, first female co-host is introduced on the 26th October 2009.

http://www.abc.net.au/tv/goodgame/video/default.htm?pres=20091026_2030&story=2 http://abc.net.au/tv/goodgame/video/vodcast/goodgame_wmv.xml

Item 40

Blog for tech Ed event Aug 26, 2010

Games Blog discussing the hiring of ‘meter maids’ as entertainment at Tech Ed event and the subsequent apology from Microsoft http://discuss.itwire.com/viewtopic.php?f=60&t=21870

Item 41

ACS Women in the Industry survey 2010

Survey questions 2010 ICT Examples of types of questions include, “what will influence your next career move?” Lists money, child minding, flexible hours. Also identifies the concept of ‘soft skills’ as non-technical personal attributes.http://www. acswsurvey.com/

Item 42

Twitter like message to Brisbane university

Dear QUT website June 2009

ICT Dear QUT is a website where students can leave feedback on their experiences at the university. “Put some compulsory IT classes in all degrees for chicks, its [sic] getting a bit lonely in the IT lectures”. http://dearqut.com/messages/81 http://discuss.itwire.com/viewtopic.php?f=60&t=21870

Item 43

Personal Blog Blogger reflects on Wired Women event

Jan 2009

Games http://zombietronics.blogspot.com/2009/01/wired-women.html Personal blog reflection of the Wired Women event

Item 44

Online university website

Technology http://news.illinois.edu/news/09/0303nerds.html American university professor discuses the stereotypes surrounding geeks. Headline “Geeks may be chic, but negative nerd stereotype still exists”

Item 45

List serv for socialissuegames

2010 Games Email announcing scholarship using language that may suggest stereotyping, for example referring to a girl as a young ‘lady’

Item 46

Press release 3/2010 Technology Presents ‘research’ that women in technology have ‘male’ brains Y

Item 47

Online news site Eweek 3/2010 Technology http://www.eweek.com/c/a/IT-Management/Report-Techs-Gender-Gap-Widened-by-Uninviting-Workplace/ Article suggests women in IT face uninviting workplaces however still enjoy their work

Item 48

Online news site Blog.wired.com

3/2010 Games In the UK games industry it is reported women earn more than male counterparts. One blog respondent suggests “It may be that women tend to work in the marketing and publishing sector of the industry, where all the money is. Men dominate the far lower paid development sector, where all the work is done.” http://blog.wired.com/games/2009/01/reportwomenin.Html

Item 49

Google search on term “Game On: Wired Women”

Jan 2009

Games Online games information site. http://www.gossipgamers.com/games-wake-up-to-girl-power/ Quotes from two women working in the games industry in Brisbane. Posted prior to Wired Women event.

Item 50

Blog Blog linked to website “tech girls are chic:not geek”.

Jan 2009

Games Blog posting from women actively involved in promoting women in technology careers. Mentions the wonderful role models working in the Brisbane gaming industry.http://techgirlsarechic.blogspot.com/2009/01/game-on-girls.html

Item 51

job advertisement Seek online job ad Dec 2010

Multimedia Web and graphic designer role. Organisational culture emphasises social aspect. Also mentions the supply of fresh fruit,

However makes a point that this is not a joke. Suggestive that eating healthy fruit is not the ‘norm’ in the industry context. Also advertised under both web and IT categories, indicative of the links to the ICT industry.

Item 52

Industry related Magazine ‘Web Designer’

Special Bulletin Women on the Web

2009 Multimedia Article interviews several female web designers working in the UK. Discusses difficulty of entering the industry.

Item 53

The Age newspaper online

Article 2008 Games http://www.theage.com.au/articles/2008/05/07/1209839660016.html?page=fullpage Article recognises more young women are playing games, however the choice of products is limited. States “Many industry insiders believe the key to creating more games that appeal to women is to get more women into the industry.”

Item 54

Insight Economics(2006) Australian Electronic Game Industry report

Report 2006 Games Report salaries for games workers. Found that in a survey of games workers there was considerable variation in salaries for some occupations. For example, salaries for software programmers ranged from $30,000-$39,999 to $90,000-$99,999. Source: http://www.edulists.com.au/yr7-10it/Resources/Industry_Profile.pdf

Item 55

Telstra FITT International Women's Day (IWD) Luncheon

Advertisement for luncheon

3/2010 ICT http://www.fitt.org.au/events.html Y

Item 56

Google search for software crack for Dreamweaver

2009 ICT/Games/Multimedia

Search for Dreamweaver software

Search date: 2007

Search date: 2009 Item 57

Website based initiative to foster women in games

American based 2011 Games

http://www.womeningamesjobs.com/

Emphasis on networking via social media.

Item 58

Website based initiative to foster women in ICT

Website, Australian founder

2011 IT Website emphasising social value of IT http://itmillion.com/

Item 59

Website- online article Online article suggesting strategies aiming to improve the ‘image’ of technology related careers

3/2010 Technology http://chronicle.com/article/Bringing-Girls-Into-the/128099/ N

Item 60

Web site promoting women’s capacity for technical skills

Website promoting women as geeks to emphasise women’s technical capabilities

3/2010 Technology Y

Item 61

Examples of female role models from games industry

Website featuring interviews with women working in games

2008 Games http://multiplayerblog.mtv.com/category/women-working-in-games/ Survey by industry organisation Sony Online Entertainment identifies women are intimated by the male majority in game development organisations.

Item 62

Technology company ‘Google’ perceived as ‘cool’ by Australian Governor General

Newspaper media article

2009 Technology Google, a key technology organisation has a public perception of being cool, which contrasts the negative connotations surrounding ‘geeks’ and technology.

Item 29 expanded

Date Archived document Jan 24 2002 The main image can be interpreted as suggesting that there are only men in the games industry and that it

is also only men who negotiate contracts.

Jan 08 2001 http://web.archive.org/web/20010108104100/http://www.gamasutra.com/ Of the ten names mentioned on the home page all would be recognised as being male names.

Feb 04 2003 http://web.archive.org/web/20030204025647/http://www.gamasutra.com/ Of the names mentioned on this page all would be recognised as being male names. One news story about games audio starts with headline 'who do the BIG BOYS turn to' [original emphasis] suggesting a male dominated industry.

Jan 10 2004 http://web.archive.org/web/20040110035122/http://www.gamasutra.com/ One year later, the same news story (see above).

Jan 01 2005 Advertising banner states, “Our passion for music & sound is matched only by our love of games”, suggesting the workforce is passionate. There are no female writers featured. Educational pathways becoming more prominent on the side bar areas.

Jan 01 2006 http://web.archive.org/web/20060101022546/http://gamasutra.com/ Side bar advert features not only young men at an event but several older women. One obviously female name mentioned on home page.

2007 http://web.archive.org/web/20070105050101/http://gamasutra.com/ Side bar features social networking site 'twitter'. One female name now as a featured writer, twelve male names.

2008 Of nine names on front page, two could possibly be seen as being either gender. There are no obviously female names anywhere on the page.

2008 http://web.archive.org/web/2008 0114013739/jobs.gamasutra.com /jobSeekerx/ViewAccountProfile. asp?accountno=24433&CPID=3139&#Profile3139 Quote on website states “We hire the best and give them the freedom to learn and do their thing, and teamwork is the norm”. This is interpreted as team work being a desired trait for the industry.

APPENDIX 18: APPROXIMATE PARTICIPATION RATES FOR WOMEN WORKING IN THE ICT, IT AND DCI (AUSTRALIA AND INTERNATIONALLY FROM 1997 TO 2007).

Table 64 Approximate participation rates for women working in the ICT, IT and DCI (Australia and internationally from 1997 to 2007). Industry Source Country Cohort Year Percentage Comments ICT industry ABS Labour Force Survey

(IBSA, 2012) AU All IT

occupations 1997-2004

25% Figures are the 2009 annual average for women’s full-time employment

VET IT training109 Newmarch, Taylor-Steele, Cumpston (2000)

AU Students in IT 1999 17%

Digital Media

Recruitment firm Direct Recruitment 107110

UK Digital Media workers

2002 11 % Compared with the boom era of 1999 to 2000 with figures over 30%.

Queensland computing professionals

Women in the Smart State 2003–2008: Directions Statement (Queensland Government 2003)

AU Computing professionals

2003 20%

American IT workforce

Untapped Talent: Diversity, Competition and America's High Tech Future (ITAA, 2005)

US IT workforce 2004 24.9% When administrative categories (Data Entry Keyers and Computer Operators) are excluded.

American IT workforce

Untapped Talent: Diversity, Competition and America's High Tech Future (ITAA, 2005)

US IT workforce 2004 32.4 % Declined from a high of 41% in 1996.

Key interactive media sectors

Skillset’s 2004 Employment Census (Skillset, 2004)

UK Employed professional

2004 33% (web) 8 % (games)

Figures are a snapshot on the day of the census

DCI (Games) IGDA survey (Gourdin, 2005) International Games industry workers

2005 11.5%

DCI (Games) Australian Bureau of Statistics (ABS)(2008a)

AU Games workers 2006-2007 financial year

10.7% Games workers employed fulltime in industry, rather than those in multimedia

Last Page

109 VET (Vocational Education and Training) 110 http://www.auroravoice.com/pressarticle.asp?articleid=100

Annex for Appendix 17

Girls Guide To Gaming

Women and Games:Support and Useful Information

www.womengamers.com

*****This site is great. It is dedicated to encouraging young women to take an interest (or to

develop a current interest) in the games industry. It communicates on a mature level to its

audience, while at the same the language is simple to understand and follow. Although

directed primarily toward women, the site is not ‘girly’ and is careful not to exclude

people from any age, experience or gender group. The site is for anyone interested in the

industry. It is easy and fun to browse, and contains an absolute wealth of information on

every subject you could possibly think of. The layout is fresh and not too busy and

dispels any myths that the industry is run by geeks. I would recommend this website to

anyone of any age who is interested in games and the gaming industry.

The site includes: Healthy interesting articles, great topical interviews, a resident

psychologist, game reviews/previews, hardware q&a, jobs, careers advice, a discussion

forum and downloads.

e.g. A free downloadable emulator and programming software that enables a child to

build their own Game Boy Advance game. It also points to a hardware add-on that makes

it possible to burn the game onto a ram card and therefore play your very own game on

your personal GBA.

www.gbadev.org

***For people who are interested in making their own game for Game Boy Advance (as

mentioned above). There is demo software that can be downloaded for free, and space to

post the game you have made along with comments so others can read about it and

download it to play on their GBAs.

This site includes: Games demo’s, sources and reference material (to use to make your

games), mailing list, comprehensive hardware reviews, competitions, forum and a help

Although the site ideally requires some prior knowledge of game building, it can be

followed easily enough by the average layman. This site is perfect for the budding game

designer/programmer.

www.sloperama.com/advice.html

*****Do NOT go past this website.

A website maintained by Tom Sloper, a versatile and prolific game

producer/designer/consultant/author and speaker who has developed games and toys

selling over 5.5 million units, written articles for books, held speeches and taught classes

in games at various high profile universities. Drawing on his extensive experience, he

offers advice and encouragement, and training on just about every subject to do with the

Games Industry. He has lessons covering subjects from Getting a Game Idea off the

Ground and Useful Study For Game Designers, to Types of Jobs in the Games

Industry, to Things You Can Do On Your Own to Prepare for a Game Career, to

Lessons on How to Get Your Foot in the Door and What to Expect from the Industry. He

even has the most comprehensive list of Game Related sites I have ever seen. There is no

topic left uncovered and this site is a must read for anyone thinking about joining, or has

a general interest in the games industry.

www.igda.org/women/

****The IGDA is the International Games Developers Association. They run a comprehensive

website www.igda.com covering all aspects of the industry. The site dedicated to women

(as listed above) was formed to create a positive impact on the game development

industry with respect to gender balance and equality. It is led by Sheri Graner Ray from

Sony Online and Heather Kelley from Ubisoft. The site includes a women_dev mailing

list, Women GDC sessions, recourse lists, reference guides and a discussion forum, and is

one of the pioneers of the girl game movement.

www.gamegirladvance.com

*****This site is a girl oriented weblog and online journal that brings alternative perspectives

to videogame culture. This site is dedicated to keeping up to date on the girl game

movement and related topics and to passing on information. The site covers important

gender issues in regards to the game industry, comprehensive discussion forums, it has a

literary journal, and shop for all your game related needs. This site is an easy way to

retrieve information and keep up abreast of the game market, and all the pros and cons

that come with it as witnessed from a female point of view, in a fun and friendly

environment.

Appendix 17

____________________________________________________________________case_item_8_Girls_Guide_To_Gaming

Newsletter Archive | Editorial Archive | Publications Archive | Other | WIG SIG Contributors

Looking for a job?Ready to take that next career step? Sign up to our LinkedIn group for occasional job postings.

To post jobs to IGDA Women in Games, sign up on our LinkedIn group and use the Jobs function. Do not post full job descriptions. Limit your postings to 3 per month.

Use the Jobs function as a job candidate by providing brief, occasional updates on your availability and current goals as a game professional, contractor, small business or student seeking an internship.

Women's participation in the Australian digital content industry

Documents

Transcript of Women's participation in the Australian digital content industry

Women's Participation in the Catholic Church in Zambia

Women's Participation, Constraints, and Opportunities for ...

SNDT Women's University

Prioritising women's safety in Australian perpetrator ...

Hearing Women's Voices?

Rebel Group Ideology and Women's Participation in Violent ...

Review of Women's Studies

COMMERCE.pdf - DUVVURU RAMANAMMA WOMEN'S ...

Women's Orgasm - CiteSeerX

On the Great Crusade: Australian participation in Normandy and the liberation of Western Europe

PATNA WOMEN'S COLLEGE

Review of High Performance Structure in Australian Women's ...

GOVERNANCE AND WOMEN'S GROUP PARTICIPATION IN ...

Geographies of Women's Health

Through the Magnifying Glass: Women's Work and Labour force participation in urban Delhi

Indigenous Women's Rights 2011

Australian Tales

PREGNANT WOMEN'S GRAPHIC REPRESENTATIONS

Books in Women's Hands

Muslim Women's Right Muslim Women's Rights Discourse in ...